探索全球前沿学术脉络

01.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2606.18293

Vibe Coding Ate My Homework: An evaluation of AI approaches to greenfield software engineering and programming

作者:

Callum Barbour↗

arXiv:2606.18293v1 Announce Type: cross Abstract: Thanks to rapid developments in generative AI, we are in the midst of a paradigm shift that may change how we interact with computers forever. We have observed a growth in the use of natural language prompts to build applications and coding infrastructures without underlying knowledge of the field, and this practice has been dubbed `vibe coding.' It arguably represents what the field of programming has been building towards since the beginning, with every higher level of abstraction that is conceived. Vibe coding promises to be the endpoint for the meta of high-level programming as far as method of input is concerned: eliminating a human's use of code syntax entirely in favour of programming in their mother tongue. This paper aims to evaluate the viability of vibe coding for greenfield software engineering tasks, as well as analyse the benchmarks that have been used to measure its software engineering prowess. To this end, we have developed an evaluation suite for analysing an LLM's proficiency in carrying out simple, isolated greenfield programming tasks in Python to provide scoped insight on the matter.

阅读与讨论 → 访问原文 →

02.

arXiv (CS.CL) 2026-06-12 DOI: arXiv:2606.07218

HKVM-RAG: Key-Value-Separated Hypergraph Evidence Organization for Multi-Hop RAG

作者:

Mingyu Zhang↗Ying Ma↗

Multi-hop RAG poses a data-engineering problem beyond passage matching: under fixed retrieval budgets, a system must organize retrieved text into evidence units that expose answer chains. Dense retrievers score passages independently, while graph-based memories make associations explicit but often rely on pairwise or entity-centered keys that fragment multi-hop evidence. We present HKVM-RAG, a key-value-separated evidence-organization layer. It assembles answer-path hyperedges from cached passage-level LLM evidence tuples and uses them as retrieval keys, while retaining passage text as answer values. To isolate key-space design, our fixed-substrate protocol holds the tuple cache, candidate passages, reader, and evaluation budget constant across pairwise graph and hypergraph variants. Weighted hypergraph key-value retrieval improves over KG-PPR by +3.426 F1 on 2WikiMultiHopQA and +3.592 F1 on MuSiQue; HotpotQA shows that higher structured support coverage need not yield standalone answer-F1 gains. We therefore study WHG-KV as an evidence-control signal rather than a dense-retrieval replacement. Oracle and train-to-dev analyses identify support selection as repairable, and a dense-aware controller combines frozen ColBERTv2 and HKVM rank/score features using out-of-fold HKVM predictions. It reaches 88.846, 65.073, and 85.810 F1 on the three benchmarks, improving over ColBERTv2 by +11.084, +6.763, and +5.966 F1. Source-level ablations show that matched non-WHG structured signals do not match the WHG-KV gains. These results provide bounded evidence that key-value-separated hypergraph organization can serve as a reusable evidence-control mechanism for multi-hop RAG.

阅读与讨论 → 访问原文 →

03.

medRxiv (Medicine) 2026-06-15 DOI: HASH:112b02b467c7f2745a8fb2b77e3d2725

Shortened blastocyst vitrification achieves live birth rates comparable to standard protocols: an analysis of 3168 cryotransfers

作者:

Ebinger↗Misurac↗Lynch↗Desai↗De Rosa↗Vassena↗Atkinson↗

Study question Do shortened blastocyst vitrification and warming protocols provide comparable live birth rates (LBR) and obstetrical and perinatal outcomes to traditional vitrification and warming protocols? Summary answer Shortened vitrification and warming protocols provide comparable LBR, obstetric and perinatal outcomes to traditional protocols. Shortened vitrification coupled with traditional multi step warming benefitted women >35yrs. What is known already Embryo viability following cryopreservation is dependent on blastomere survival and functional integrity, both impacted by ice crystal formation and osmotic gradients. Recent innovations in cryopreservation challenge the need for stepwise dehydration and rehydration protocols. While one step ''fast'' blastocyst warming protocols seem to provide equivalent clinical outcomes to traditional ''slow'' protocols, fewer studies investigate whether blastocyst dehydration rates can be similarly increased. A thorough safety and effectiveness evaluation remains necessary for both treatment success and offspring health. Study design, size, duration Three clinics within a network participated in this retrospective consecutive cohort study, with cycle data collected for 3603 warmed blastocysts resulting in 3168 frozen blastocyst transfers in 2170 patients between 2023 and 2025. We modelled the relationship between ''fast'' versus ''slow'' protocols and outcomes with Generalized Additive Models, and linear and logistic regressions where appropriate. Two tailed chi square with Yates correction was used to examine pregnancy loss and obstetrical and perinatal outcomes; p0.05). Importantly, women 35yrs or older at vitrification (n=1715 transfers) profited from a F/S strategy, which provided a significant increase in live birth rates (OR:1.42 [1.02-1.98] p=0.038) compared to S/S. The same improved live birth following a F/S strategy were also seen in embryos of lower quality (OR:1.78 [1.12-2.83] p=0.015), suggesting of a protective effect of this cryopreservation strategy on the developmental competence of impaired germplasm. Limitations, reasons for caution Factors affecting the results may be unaccounted for by the study retrospective nature. Wider implication of the findings Overall, shortened, ''faster'' vitrification and warming protocols provide comparable reproductive outcomes to traditional ones. The combination of shorter exposure to cryoprotectant (CPA) during vitrification and stepwise osmotic gradient during warming provided significant clinical benefits specifically to patients >35 and lower quality embryos, pointing to the possibility of adapting vitrification protocols to specific patients populations and optimizing their clinical outcomes.

阅读与讨论 → 访问原文 →

04.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2606.01139

SkillRevise: Improving LLM-Authored Agent Skills via Trace-Conditioned Skill Revision

作者:

Yuxuan Liu↗Zhaochen Su↗Lingyun Xie↗Yuhao Zhang↗Qing Zong↗Jiahe Guo↗Zhongwei Xie↗Yiyan Ji↗Yauwai Yim↗Hongyu Luo↗Xiyu Ren↗Ruan Chenyu↗…

arXiv:2606.01139v3 Announce Type: replace Abstract: Agent skills are procedural artifacts that enable LLM agents to execute workflows, verify constraints, and recover from failures. Existing self-evolving methods refine skills using accumulated trajectories. However, they struggle in cold-start settings, where only an initial, imperfect skill is available. Consequently, skill construction defaults to expert authoring or one-shot LLM generation. Expert-authored skills are costly and may not align with how LLM agents actually execute tasks, while one-shot generated skills can be syntactically well formed yet behaviorally weak. To bridge this gap, we propose SkillRevise, an execution-grounded framework designed to iteratively refine these initial skills. SkillRevise diagnoses skill defects from execution evidence, retrieves relevant repair principles from a general memory, and applies execution-anchored edits. By re-executing candidates, it retains the first verifier-passing skill within the revision budget and falls back to empirical utility only when no candidate succeeds. Evaluated across three benchmarks and five LLMs, SkillRevise substantially outperforms one-shot baselines, improving the base agent's success rate on SkillsBench from 36.05% to 61.63%. Furthermore, the revised skills transfer across both executors and task environments, suggesting that SkillRevise captures reusable procedural knowledge beyond any single executor.

阅读与讨论 → 访问原文 →

05.

PLOS Computational Biology 2026-06-02 DOI: HASH:bd6be80d4241cffa285aba644288cace

Assessing the importance of sex and disease-specific anatomy in electrophysiology and mechanical simulations with a newly developed public virtual cohort of four-chamber heart models

作者:

José Alonso Solís-Lemus↗

by José Alonso Solís-Lemus, Rosie K. Barrows, Cristobal Rodero, Marina Strocchi, Natalie Montarello, Nishant Lahoti, Cesare Corrado, Abdul Qayyum, Shahrokh Rahmani, Caroline Roney, Gernot Plank, Christoph Augustin, Hao Xu, Alistair Young, Pras Pathmanathan, Ronak Rajani, Steven A. Niederer This work presents a study on how differences in cardiac anatomy attributed to sex and disease can influence cardiac electrophysiology and mechanics using a virtual cohort of four-chamber heart models. Patient anatomy varies across sex and disease. However, capturing this variation in in-silico studies remains poorly accounted for, with studies often using either single representative cases or imbalanced virtual cohorts. Whole-heart electromechanics models incorporate the patient’s anatomy, electrophysiology and mechanics across different scales, from molecular, tissue and whole-heart and circulatory system levels. However, cardiac models are typically built from one or a small number of anatomies, with sex rarely reported and the effects of anatomical variability, which include those due to sex or disease, largely unexplored. This limits clinical translation and reduces regulatory credibility. We developed fifty patient-specific anatomical models of 25 male and 25 female hearts in heart failure and control cases. We ran benchmark passive inflation and paced activation simulations with consistent parameters and boundary conditions across cases to isolate the impact of anatomical variations with sex and disease. Heart failure models exhibited increased chamber volumes, larger volume changes during inflation, and delayed activation times relative to controls. These trends were consistent across sexes, although right ventricular activation showed a significant sex-based difference. Variations in anatomy with sex and disease have a significant impact on cardiac simulations, which support the inclusion of multiple heart anatomical models in in-silico trials. The resulting virtual cohort captures key anatomical variability and is publicly available, along with the underlying code (see Data Availability statement).

阅读与讨论 → 访问原文 →

06.

arXiv (CS.CV) 2026-06-18 DOI: arXiv:2606.18886

DINO-Med3D: Bridging Dimension and Domain Gaps in Volumetric Segmentation via Progressive Adaptation

作者:

Haoyu Hu↗Xiyao Ma↗Shiqi Liu↗Linsen Zhang↗Xiaoliang Xie↗Xiaohu Zhou↗Zeng-Guang Hou↗

Although DINOv3 has demonstrated remarkable semantic discrimination in natural imagery, its direct application to volumetric medical segmentation is hindered by inherent dimension and domain disparities. To resolve these issues, we propose DINO-Med3D, a two-stage progressive framework that repurpose the pre-trained DINOv3 encoder for 3D medical tasks. In the first stage, we mitigate the dimension gap by introducing a multi-slice embedding module that incorporates pseudo-3D context, while simultaneously employing a segmentation proxy task to adapt representations learned from natural scenes to the medical domain. Subsequently, we further enhance volumetric understanding by adding lightweight 3D adapters into the frozen backbone to enforce global inter-slice continuity. Finally, to compensate for the spatial information loss inherent in the embedding process, we design a parallel detail recovery stream to explicitly preserve high-frequency boundary cues. Extensive experiments on five public datasets demonstrate that our approach successfully adapts DINOv3 to the medical domain and significantly outperforms state-of-the-art baselines.

阅读与讨论 → 访问原文 →

07.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2512.22827

FasterPy: An LLM-based Code Execution Efficiency Optimization Framework

作者:

Yue Wu↗Minghao Han↗Ruiyin Li↗Peng Liang↗Amjed Tahir↗Zengyang Li↗Qiong Feng↗Mojtaba Shahin↗

arXiv:2512.22827v2 Announce Type: replace-cross Abstract: Code often suffers from performance bugs. These bugs necessitate the research and practice of code optimization. Traditional rule-based methods rely on manually designing and maintaining rules for specific performance bugs (e.g., redundant loops, repeated computations), making them labor-intensive and limited in applicability. In recent years, machine learning and deep learning-based methods have emerged as promising alternatives by learning optimization heuristics from annotated code corpora and performance measurements. However, these approaches usually depend on specific program representations and meticulously crafted training datasets, making them costly to develop and difficult to scale. With the booming of Large Language Models (LLMs), their remarkable capabilities in code generation have opened new avenues for automated code optimization. In this work, we proposed FasterPy, a low-cost and efficient framework that adapts LLMs to optimize the execution efficiency of Python code. FasterPy combines Retrieval-Augmented Generation (RAG), supported by a knowledge base constructed from existing performance-improving code pairs and corresponding performance measurements, with Low-Rank Adaptation (LoRA) to enhance code optimization performance. Our experimental results on the Performance Improving Code Edits (PIE) benchmark demonstrate that our method outperforms existing models on multiple metrics. The FasterPy tool and the experimental results are available at https://github.com/WuYue22/fasterpy.

阅读与讨论 → 访问原文 →

08.

arXiv (CS.AI) 2026-06-11 DOI: arXiv:2606.11769

When Do Data-Driven Systems Exhibit the Capability to Infer?

作者:

Maximilian Poretschkin↗Tabea Naeven↗

arXiv:2606.11769v1 Announce Type: new Abstract: The European AI Act is the first comprehensive regulation of artificial intelligence (AI), setting out extensive obligations, particularly for so-called high-risk and general-purpose AI systems. A key distinguishing feature of AI systems under the AI Act is the capability to infer. Since the AI Act does not clearly define what inference is, there is a gray area for certain data-driven systems. A specific example is credit scoring systems, which are listed by Annex III of the AI Act. At the same time, however, these are often implemented using statistical models for which it is unclear whether they have the capability to infer and thus fall under the AI definition of the AI Act at all. Motivated by statistical learning theory, this work develops a framework for grading different levels of the capability to infer. Based on the AI Act and the Commission Guidelines on the definition of an artificial intelligence system, we analyze which levels constitute sufficient capability to infer within the meaning of the AI Act and where further regulatory clarity is needed. We illustrate the framework by creating two realistic credit scoring workflows and show whether and where inference occurs in them. Our analysis illustrates that not only individual models but the entire data processing workflow must be considered. It also shows that the involvement of human experts during development can have significant influence on the capability to infer. Code can be found at https://github.com/fraunhofer-iais/inference-framework-creditscorecards.

阅读与讨论 → 访问原文 →

09.

arXiv (CS.CV) 2026-06-12 DOI: arXiv:2606.13461

Reinforcement Learning for Neural Model Editing

作者:

Shaivi Malik↗

Editing pretrained neural networks requires specialized algorithms tailored to specific objectives. Designing such algorithms is often time-consuming and demands significant effort. We present an exploratory framework that formulates neural model editing as a reinforcement learning problem, where agents modify models using reward feedback. We introduce two environments: MaskWorld, where agents scale weights multiplicatively, and ShiftWorld, where agents apply additive weight updates. The reward function combines a utility-preservation objective with a task-specific editing objective, enabling agents to learn targeted modifications while maintaining overall model performance. We evaluate the framework on bias mitigation in text classification and machine unlearning in image classification, both of which traditionally rely on specialized algorithms. Our results show that the learned policies reduce forget set accuracy to nearly 0% while preserving over 90% retain set accuracy on the unlearning task. In the bias mitigation setting, the learned policies improve bias-related performance by more than 5% while maintaining general classification utility. Our findings show that neural model editing can be cast as a reinforcement learning problem, allowing editing policies to be learned from reward feedback rather than manually engineered for each task.

阅读与讨论 → 访问原文 →

10.

arXiv (CS.AI) 2026-06-12 DOI: arXiv:2606.12647

Token Complexity Theory for AI-Augmented Computing

作者:

Jie Wang↗

arXiv:2606.12647v1 Announce Type: cross Abstract: AI-augmented computing delegates natural language queries, code generation requests, and other open-ended tasks to a cluster of AI models that processes queries and generates responses. This paradigm introduces a resource dimension that neither classical time nor space complexity captures: the cost of sending queries to and receiving responses from such a cluster. We introduce token complexity, a formal resource measure defined as the minimum expected token cost to achieve a specified level of output quality on a task, and develop a taxonomy classifying AI systems by the strength of their probabilistic properties. We develop token complexity within the framework of AI-Oracle Turing machines, in which a probabilistic Turing machine interacts with a stochastic oracle via dedicated query and response tapes. We prove basic theorems establishing that token complexity behaves as expected: monotonicity (higher quality costs more tokens), convexity (quality improvements become progressively more expensive), price sensitivity (small price changes produce bounded cost changes), and price-relativity of task ordering (the token complexity ordering of tasks can reverse depending on the query-to-response cost ratio). We prove that the complexity frontier, defined as the set of all feasible resource bounds in tokens, time, and space, is non-empty, upward-closed, and convex.

阅读与讨论 → 访问原文 →

11.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.15589

Is Code Better Than Language for Algorithmic Reasoning

作者:

Terry Tong↗Yu Feng↗Surbhi Goel↗Dan Roth↗

arXiv:2606.15589v1 Announce Type: cross Abstract: For tool-augmented language models, comparing natural-language reasoning with code-execution pipelines is difficult because the comparison changes both the intermediate representation and the execution mechanism. We separate these factors with an intermediate intervention: the model expresses its reasoning as executable code, and the language model simulates that code in context to produce an answer. On a 40-task verifiable algorithmic benchmark, deterministic code execution outperforms natural-language reasoning by +31.6pp. We observe that the intermediate intervention is not meaningfully different from natural-language reasoning (+0.15pp). These results suggest that, in our evaluated setting, changing the intermediate representation alone does not explain the tool-use advantage, providing evidence for the performance gains requiring reliable external execution. We formalize this intuition with a simple statistical decision-theoretic model that characterizes when execution dominates end-to-end risk in our disentangled trace-generation/execution regime. We validate our theory using a reconstruction intervention that leverages a proxy language model to infer natural-language reasoning traces from code representations, recovering performance comparable to the original natural-language reasoning pipeline. All experiments are at https://github.com/TerryTong-Git/ToolProj.

阅读与讨论 → 访问原文 →

12.

arXiv (CS.CL) 2026-06-11 DOI: arXiv:2601.07506

Judging Against the Reference: Uncovering Knowledge-Driven Failures in LLM-Judges on QA Evaluation

作者:

Dongryeol Lee↗Yerin Hwang↗Taegwan Kang↗Minwoo Lee↗Younhyung Chae↗Kyomin Jung↗

While large language models (LLMs) are increasingly used as automatic judges for question answering (QA) and other reference-conditioned evaluation tasks, little is known about their ability to adhere to a provided reference. We identify a critical failure mode of such reference-based LLM QA evaluation: when the provided reference conflicts with the judge model's parametric knowledge, the resulting scores become unreliable, substantially degrading evaluation fidelity. To study this phenomenon systematically, we introduce a controlled swapped-reference QA framework that induces reference-belief conflicts. Specifically, we replace the reference answer with an incorrect entity and construct diverse pairings of original and swapped references with correspondingly aligned candidate answers. Surprisingly, grading reliability drops sharply under swapped references across a broad set of judge models. We empirically show that this vulnerability is driven by judges' over-reliance on parametric knowledge, leading judges to disregard the given reference under conflict. Finally, we find that this failure persists under common prompt-based mitigation strategies, highlighting a fundamental limitation of LLM-as-a-judge evaluation and motivating reference-based protocols that enforce stronger adherence to the provided reference.

阅读与讨论 → 访问原文 →

13.

arXiv (CS.LG) 2026-06-16 DOI: arXiv:2511.09465

Branching Flows: Discrete, Continuous, and Manifold Flow Matching with Splits and Deletions

作者:

Lukas Billera↗Hedwig Nora Nordlinder↗Jack Collier Ryder↗Anton Oresten↗Aron St{\aa}lmarck↗Theodor Mosetti Bj\"ork↗Ben Murrell↗

arXiv:2511.09465v4 Announce Type: replace-cross Abstract: Diffusion and flow matching approaches to generative modeling have shown promise in domains where the state space is continuous, such as image generation or protein folding & design, and discrete, exemplified by diffusion large language models. They offer a natural fit when the number of elements in a state is fixed in advance (e.g. images), but require ad hoc solutions when, for example, the length of a response from a large language model, or the number of amino acids in a protein chain is not known a priori. Here we propose Branching Flows, a generative modeling framework that, like diffusion and flow matching approaches, transports a simple distribution to the data distribution. But in Branching Flows, the elements in the state evolve over a forest of binary trees, branching and dying stochastically with rates that are learned by the model. This allows the model to control, during generation, the number of elements in the sequence. We also show that Branching Flows can compose with any flow matching base process on discrete sets, continuous Euclidean spaces, smooth manifolds, and `multimodal' product spaces that mix these components. We demonstrate this in three domains: small molecule generation (multimodal), antibody sequence generation (discrete), and protein backbone generation (multimodal), and show that Branching Flows is a capable distribution learner with a stable learning objective, and that it enables new capabilities.

阅读与讨论 → 访问原文 →

14.

arXiv (CS.CL) 2026-06-15 DOI: arXiv:2505.04671

Reward-SQL: Boosting Text-to-SQL via Stepwise Execution-Aware Reasoning and Process-Supervised Rewards

作者:

Yuxin Zhang↗Meihao Fan↗Ju Fan↗Mingyang Yi↗Yuyu Luo↗Guoliang Li↗Bin Wu↗Wenchao Zhou↗

Recent advances in large language models (LLMs) trained with reinforcement learning (RL) have improved Text-to-SQL performance. However, RL-based approaches still struggle with complex queries due to two key limitations: insufficient stepwise execution-aware reasoning grounded in database feedback, and the lack of process-level rewards for guiding reasoning optimization. To address these issues, we propose CoCTE, a divide-and-conquer and execution-aware reasoning framework that progressively composes SQL queries through intermediate view validation and structured Common Table Expressions (CTEs), improving both accuracy and interpretability. To realize a CoCTE reasoning process, we develop Reward-SQL, a unified approach with three stages: (1) model initialization, which equips LLMs with structured CoCTE reasoning capabilities; (2) process reward design, which delivers fine-grained, execution-aware supervision; and (3) process-supervised RL and inference, which integrates process rewards into training and guides the inference stage by process rewards. This paper addresses the core challenges in Reward-SQL and makes the following contributions. We introduce a process reward model (PRM) that combines execution-aware trajectory scoring with entropy-based step weighting, providing dense and interpretable supervision across reasoning steps. We integrate PRM into both RL training and inference stages, stabilizing optimization and improving trajectory exploration with process-level signals. Experiments show that Reward-SQL significantly outperforms baselines with comparable model sizes, and exhibits strong cross-domain generalization.

阅读与讨论 → 访问原文 →

15.

arXiv (CS.AI) 2026-06-19 DOI: arXiv:2504.11171

TerraMind: Large-Scale Generative Multimodality for Earth Observation

作者:

Johannes Jakubik↗Felix Yang↗Benedikt Blumenstiel↗Erik Scheurer↗Rocco Sedona↗Stefano Maurogiovanni↗Jente Bosmans↗Nikolaos Dionelis↗Valerio Marsocci↗Niklas Kopp↗Rahul Ramachandran↗Paolo Fraccaro↗…

arXiv:2504.11171v5 Announce Type: replace-cross Abstract: We present TerraMind, the first any-to-any generative, multimodal foundation model for Earth observation (EO). Unlike other multimodal models, TerraMind is pretrained on dual-scale representations combining both token-level and pixel-level data across modalities. On a token level, TerraMind encodes high-level contextual information to learn cross-modal relationships, while on a pixel level, TerraMind leverages fine-grained representations to capture critical spatial nuances. We pretrained TerraMind on nine geospatial modalities of a global, large-scale dataset. In this paper, we demonstrate that (i) TerraMind's dual-scale early fusion approach unlocks a range of zero-shot and few-shot applications for Earth observation, (ii) TerraMind introduces "Thinking-in-Modalities" (TiM) – the capability of generating additional artificial data during finetuning and inference to improve the model output – and (iii) TerraMind achieves beyond state-of-the-art performance in community-standard benchmarks for EO like PANGAEA. The pretraining dataset, the model weights, and our code are open-sourced under a permissive license.

阅读与讨论 → 访问原文 →

16.

arXiv (CS.AI) 2026-06-17 DOI: arXiv:2606.17897

Learn to Quantify Social Interaction with Constraints for Pedestrian Walking

作者:

Xiaodan Shi↗

arXiv:2606.17897v1 Announce Type: new Abstract: Long-term human path forecasting in crowds is critical for autonomous moving platforms (like autonomous driving cars and social robots) to avoid collision and make high-quality planning. Although the current research take into account social interactions for prediction, they don't reveal the exact kinds of social interactions happened among people and how the social interactions affect the decision-making process of pedestrians, which further limits its robustness. Social interactions in pedestrian walking are intuitively massive and hard to label and quantify. In this paper, we explore creatively to quantify and interpret how pedestrians interact with others by proposing Learn to Cluster. Our clustering social interactions is probabilistic latent variable generative, learning directly from sequential trajectory observations, scalable to arbitrary number of pedestrians. Learn to cluster is label-free and can be naturally integrated into the training process of the prediction model. The latent variables will then serve as 'labels' to categorize social interactions. Extensive experiments over several trajectory prediction benchmarks demonstrate that our method is able to learn the patterns of social interactions and effectively integrate the patterns to pedestrian trajectory prediction.

阅读与讨论 → 访问原文 →

17.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2606.10740

When the Chain of Thought Knows Better: Failure Modes in Multi-Turn Reasoning Models

作者:

Sai Kartheek Reddy Kasu↗Nils Lukas↗Samuele Poppi↗

Failures in multi-turn reasoning models are largely invisible to terminal-score evaluation. A model can lock onto an unsafe stance early in a long dialogue, yet its final-turn refusal rate may appear indistinguishable from a robustly aligned baseline. To expose these hidden temporal dynamics, we propose a trace-level diagnostic - the CoT-Output 2x2 safety matrix. This framework labels every turn along two independent axes (internal reasoning and visible output), yielding four operationally defined failure cells: robust alignment, alignment faking, overt jailbreak, and a distinct failure mode we term context-injection failure (where the CoT maintains safe reasoning, but the visible output produces harm, highlighting a multi-turn manifestation of reasoning unfaithfulness). We evaluate three distilled reasoning targets against a fixed attacker across five oversight conditions, collecting 6750 turn-level observations on the Information-Hazard scenario. Our analysis reveals two reproducible vulnerabilities: an oversight paradox where explicit monitoring cues paradoxically increase alignment-faking rates rather than suppress them, and a context-injection failure where models lock onto unsafe external outputs despite safe internal states. We release the full dataset of multi-turn dialogues and CoT traces to support follow-up trace-diagnostic research.

阅读与讨论 → 访问原文 →

18.

bioRxiv (Bioinfo) 2026-06-18 DOI: HASH:b5f7b3d5762651501d1e794d94daa51e

Bioinf-Farma: supervised integration of epitope prediction and recombinant protein developability for automated vaccine candidate prioritization

作者:

Bondi↗Crespi↗Orlando↗Lescai↗Serapian↗S. A↗Colombo↗Fasano↗Pollegioni↗Molla↗

Vaccine antigen discovery requires prioritizing protein candidates according to both immunogenic potential and recombinant expression feasibility. These properties are typically evaluated using separate computational tools, requiring researchers to integrate heterogeneous outputs through ad hoc workflows. Here, we present BIOINF-farma, a modular platform integrating epitope prediction and developability assessment for rational antigen selection within a unified environment. Candidates can be submitted as amino acid sequences or three-dimensional structures. When experimental structures are unavailable, BIOINF-farma automatically searches for models in AlphaFold DB or performs structure prediction using Boltz-2, ensuring a standardized structural representation for downstream analyses. Antigenicity is quantified by combining structure-based conformational epitope signals (MLCE/REBELOT-BEPPE) and sequence-based linear epitope propensity scores (BepiPred 3.0) into a protein-level Antigenicity Score, with a classification threshold optimized on a manually curated validation dataset. Developability is evaluated through two supervised Random Forest meta-learners that integrate three solubility predictors (DeepSoluE, SoluProt, Protein-Sol) and three thermal stability predictors (TemStaPro, ProLaTherm, BertThermo), whose outputs are combined into an Expression Efficiency Score (EES). By integrating complementary predictive signals, the meta-learning framework achieves greater accuracy and robustness than individual predictors while maintaining performance across a broad range of sequence identities. The Antigenicity Score effectively discriminates antigenic from non-antigenic proteins with a large effect size, whereas EES successfully distinguishes soluble from insoluble outcomes on an independent panel of recombinant proteins expressed in Escherichia coli. BIOINF-farma jointly assesses antigenicity and expression feasibility within a single framework. Its modular architecture facilitates the incorporation of future predictive methods, while its web-based interface makes the full pipeline accessible to users without programming expertise, supporting rapid candidate triage in vaccine research and emerging pathogen responses.

阅读与讨论 → 访问原文 →

19.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2602.08597

An Attention Mechanism for Robust Multimodal Integration in a Global Workspace Architecture

作者:

Roland Bertin-Johannet↗Lara Scipio↗Leopold Mayti\'e↗Rufin VanRullen↗

arXiv:2602.08597v3 Announce Type: replace Abstract: Robust multimodal systems must remain effective when some modalities are noisy, degraded, or unreliable. Existing multimodal fusion methods often learn modality selection jointly with representation learning, making it difficult to determine whether robustness comes from the selector itself or from full end-to-end co-adaptation. Motivated by Global Workspace Theory (GWT), we study this question using a lightweight top-down modality selector operating on top of a frozen multimodal global workspace. We evaluate our method on two multimodal datasets of increasing complexity: Simple Shapes and MM-IMDb 1.0, under structured modality corruptions. The selector improves robustness while using far fewer trainable parameters than end-to-end attention baselines, and the learned selection strategy transfers better across downstream tasks, corruption regimes, and even to a previously unseen modality. Beyond explicit corruption settings, on the MM-IMDb 1.0 benchmark, we show that the same mechanism improves the global workspace over its no-attention counterpart and yields decent benchmark performance.

阅读与讨论 → 访问原文 →

20.

medRxiv (Medicine) 2026-06-11 DOI: HASH:ecd7faf38c297e4ccf09cebdb194413c

Dissecting the functional landscape of rare diseases through genomic variation in a heterogeneous cohort of 11,000 patients

作者:

Uria-Regojo↗Fernandez-Caballero↗Lopez-Alcojor↗Lopez-Lopez↗Benitez↗Rodilla↗Avila Fernandez↗Trujillo-Tiebas↗M. J↗Osorio↗Corton↗Almoguera↗…

Rare diseases (RDs) remain a major diagnostic challenge. Genetic and phenotypic heterogeneity, incomplete knowledge of disease mechanisms, and limitations in variant clinical interpretation leave many patients without a molecular diagnosis. Meanwhile, the growing volume of genomic data generated in clinical practice offers an opportunity to develop data-driven methodologies for exploring disease mechanisms and improving the reanalysis of unsolved cases. We aggregated real-world genomic data from 11,084 unrelated patients with suspected RD. Patients were clinically classified into 122 diseases. We built a multi-disease genomic variant frequency database (FJD-DB), which enabled the development of variant and gene-disease association scores by means of case-control subcohort comparisons across 32 disease groups. Functional enrichment analyses were then used to highlight disease-associated protein domains, pathways, biological processes, and phenotypes. Finally, the resulting knowledge was integrated into a data-driven framework for the guided reanalysis of unsolved RD patients applied to Inherited Retinal Dystrophies (IRD) patients as first use case. FJD-DB contained more than 45 million unique variants, including ~185,000 potentially pathogenic variants. Disease-specific analyses identified disease-associated pathogenic variants and highlighted both established and candidate disease genes. We detected 179 significantly enriched protein domains across 23 diseases, 124 Human Phenotype Ontology terms across 13 diseases, 79 Reactome pathways across 10 diseases, and 72 Gene Ontology biological processes across 8 diseases, revealing highly disease-specific functional signatures. Integration of disease-specific variant, gene, and functional association signals enabled the development of a data-driven framework for guided reanalysis of unsolved RD cases. Applied to more than 1,100 unsolved IRD cases, the framework generated clinically relevant findings in 26 patients, including four molecular diagnoses, seven candidate diagnoses, and 15 cases upgraded from non-informative findings to variants of uncertain significance. Aggregated real-world genomic data can be leveraged to identify disease-associated molecular signals generating novel biological hypotheses. A unified analytical framework provides a scalable strategy for knowledge discovery and guided reanalysis, facilitating the identification of overlooked and potentially novel genetic causes of RDs.

阅读与讨论 → 访问原文 →

21.

arXiv (CS.AI) 2026-06-19 DOI: arXiv:2606.19528

Techniques for Peak Memory Reduction for LoRA Fine-tuning of LLMs on Edge Devices

作者:

Hassan Dbouk↗Matthias Reisser↗Prathamesh Mandke↗Likhita Arun Navali↗Christos Louizos↗

arXiv:2606.19528v1 Announce Type: cross Abstract: Fine-tuning of Large Language Models (LLMs) using Low-Rank Adaptation (LoRA) on an end-user's data offers personalized experiences while keeping data private, but faces severe memory constraints on consumer hardware. Peak memory during fine-tuning often exceeds device limits, especially for models with billions of parameters and long-context training data. This paper introduces a suite of complementary techniques to reduce memory footprint without sacrificing model quality: (1) base model quantization with on-the-fly dequantization, (2) memory-efficient checkpointing combining selective activation caching and disk offloading, (3) softmax approximation using semantically relevant token subsets, and (4) logits masking. Experiments on Llama-3.2 3B and Qwen-2.5 3B demonstrate up to $26\times$ and $28\times$ reduction in peak memory, enabling fine-tuning on resource-constrained devices.

阅读与讨论 → 访问原文 →

22.

arXiv (CS.CL) 2026-06-12 DOI: arXiv:2604.10389

BLUEmed: Retrieval-Augmented Multi-Agent Debate for Clinical Error Detection

作者:

Saukun Thika You↗Nguyen Anh Khoa Tran↗Wesley K. Marizane↗Hanshu Rao↗Qiunan Zhang↗Xiaolei Huang↗

Terminology substitution errors in clinical notes, where one medical term is replaced by a linguistically valid but clinically different term, pose a persistent challenge for automated error detection in healthcare. We introduce BLUEmed, a multi-agent debate framework augmented with hybrid Retrieval-Augmented Generation (RAG) that combines evidence-grounded reasoning with multi-perspective verification for clinical error detection. BLUEmed decomposes each clinical note into focused sub-queries, retrieves source-partitioned evidence through dense, sparse, and online retrieval, and assigns two domain expert agents distinct knowledge bases to produce independent analyses; when the experts disagree, a structured counter-argumentation round and cross-source adjudication resolve the conflict, followed by a cascading safety layer that filters common false-positive patterns. We evaluate BLUEmed on a clinical terminology substitution detection benchmark under both zero-shot and few-shot prompting with multiple backbone models spanning proprietary and open-source families. Experimental results show that BLUEmed achieves the best accuracy (69.13%), ROC-AUC (74.45%), and PR-AUC (72.44%) under few-shot prompting, outperforming both single-agent RAG and debate-only baselines. Further analyses across six backbone models and two prompting strategies confirm that retrieval augmentation and structured debate are complementary, and that the framework benefits most from models with sufficient instruction-following and clinical language understanding.

阅读与讨论 → 访问原文 →

23.

arXiv (quant-ph) 2026-06-16 DOI: arXiv:2605.05167

Entanglement-Rank Duality in Quadratic Phase Quantum States

作者:

Zakaria Dahbi↗Amelle Za\"ir↗

arXiv:2605.05167v2 Announce Type: replace Abstract: Absolutely maximally entangled (AME) states are fundamental resources in quantum information theory, yet their construction and certification remain a nontrivial problem. Within the family of quadratic phase quantum states, defined by symmetric matrices $P$ over finite fields $\mathbb{F}_{p^m}$, we show that the Rank-Purity Duality $\operatorname{Tr}(\rho_S^2) = |\mathbb{F}|^{-\operatorname{rk}_{\mathbb{F}}(P_{S,\bar{S}})}$ follows from additive character orthogonality and holds over all $\mathbb{F}_{p^m}$, yielding a polynomial-time AME certification criterion. For square-free dimensions $d = p_1\cdots p_r$, the Chinese Remainder Theorem induces a prime-field factorisation. This implies additivity of Rényi-2 entropy and yields sharp obstruction criteria that rule out cases such as $\operatorname{AME}(4,6)$ and constrain the open case $\operatorname{AME}(8,6)$. As a proof of concept, we construct an explicit $\operatorname{AME}(17,10001)$ state, certified across all $65{,}535$ bipartitions, demonstrating that the framework scales to large systems and previously inaccessible local dimensions.

阅读与讨论 → 访问原文 →

24.

arXiv (CS.CV) 2026-06-16 DOI: arXiv:2606.15305

CoMNeT: A MedNeXt-CorrDiff Framework for Volumetric Brain Tumor Segmentation

作者:

Michael L. Evans↗MD Fayaz Bin Hossen↗MD Shibly Sadique↗Walia Farzana↗Khan M. Iftekharuddin↗

Accurate brain tumor segmentation from multiparametric magnetic resonance imaging (MRI) is critical for treatment planning, response assessment, and quantitative neuro-oncology research. However, automated segmentation remains a difficult task in computer vision because of variation in tumor appearance and MRI protocols across patient scans. Moreover, clinically important regions such as enhancing tumor (ET) and tumor core (TC) are often small relative to the full brain volume, furthering increasing the difficulty of achieving high voxel-level precision. In this paper, we show that combining a modern 3D convolutional segmentation model with corrective diffusion-based refinement and ensembling improves volumetric glioma segmentation on the UTSW-Glioma dataset. We propose CoMNeT, a MedNeXt-CorrDiff framework that uses four MRI modalities as input and predicts ET, TC, and whole tumor (WT) regions for automated brain tumor segmentation. MedNeXt is used as the primary segmentation model with Global Response Normalization for feature learning, while CorrDiff is trained as a postprocessing residual refinement method to correct errors in the probability maps before final thresholding. Using five-fold cross-validation, CoMNeT achieved the highest Dice score for most tumor regions, with ET, TC, WT, and average Dice scores of 0.7543 +/- 0.0261, 0.6806 +/- 0.0166, 0.9049 +/- 0.0128, and 0.7798 +/- 0.0184, respectively. CoMNeT outperformed two selected baseline models: SegResNet (0.7555 +/- 0.0190 average Dice) and standalone MedNeXt (0.7697 +/- 0.0154 average Dice). Our findings support the use of corrective diffusion and fold-level probability ensembling as practical additions to existing state-of-the-art 3D convolutional models for automated glioma segmentation.

阅读与讨论 → 访问原文 →

25.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2606.18936

SciRisk-Bench: A Risk-Dimension-Aware Benchmark for AI4Science Safety

作者:

Linghao Feng↗Yinqian Sun↗Dongqi Liang↗Sicheng Shen↗Chenfei Yan↗Yuxuan Peng↗Yilin Zhao↗Haibo Tong↗Kai Li↗FeiFei Zhao↗Yi Zeng↗

arXiv:2606.18936v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly embedded in AI for Science (AI4Science) workflows, from scientific question answering and literature analysis to laboratory planning and autonomous discovery. This progress creates an urgent need for safety benchmarks that evaluate not only scientific competence, but also whether models recognize and avoid risks in high-stakes scientific contexts. Existing AI4Science safety datasets cover several disciplines and task formats, leaving the underlying risk dimensions underspecified. We introduce SciRisk-Bench, a benchmark designed to evaluate AI4Science safety from two complementary perspectives: explicit risk dimensions and scientific disciplines. SciRisk-Bench covers 7 disciplines, 31 subdisciplines and 10 risk dimensions. In the experimental section, we evaluate both mainstream LLMs and science-oriented LLMs across risk dimensions, disciplines, and sub-disciplines, enabling fine-grained diagnosis of where scientific models remain unsafe.

阅读与讨论 → 访问原文 →