探索全球前沿学术脉络

01.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.15653

IoT-Zoo: A Container-Based Framework for Heterogeneous IoT Device Profiles and Reproducible Traffic Capture

作者:

Vagner E. Quincozes↗Diego Kreutz↗Silvio E. Quincozes↗

arXiv:2606.15653v1 Announce Type: cross Abstract: The validation of networking and security solutions for the Internet of Things (IoT) requires realistic and reproducible experimental data. However, existing platforms often achieve scalability by replicating a limited set of device types, which restricts profile diversity and fails to capture the heterogeneity of real-world IoT environments. In this paper, we present IoT-Zoo, a container-based testbed designed to support reproducible experimentation through heterogeneous, dataset-driven IoT device profiles. Built upon Containernet, IoT-Zoo automates the deployment of multi-domain scenarios and supports real application protocols such as MQTT and RTSP. The platform provides a single-command interface for environment provisioning and automated traffic capture (PCAP), enabling the generation of consistent traffic baselines and reducing the operational effort required to evaluate networking and security solutions.

阅读与讨论 → 访问原文 →

02.

arXiv (CS.LG) 2026-06-11 DOI: arXiv:2606.11652

IAPO: Input Attribution-Aware Policy Optimization for Tool Use in Small Multimodal Agents

作者:

Yifan Yang↗Zhen Zhang↗Jiayi Tian↗Liyan Tan↗Zheng Zhang↗

arXiv:2606.11652v1 Announce Type: new Abstract: This paper investigates reinforcement learning (RL) methods for improving tool-calling capabilities in multimodal small language model (SLM) agents. While existing works have explored various reward designs to improve agentic tool-calling ability, these approaches face inherent limitations for SLM training, especially under multimodal scenarios. First, many existing methods evaluate tool use correctness through exact matching against certain ground-truth or predefined formats. However, this assumption is often unsuitable for multimodal tasks, where multiple tool use paths may be valid and annotated tool trajectories are typically unavailable. Second, such sparse and brittle binary rewards provide little guidance on how to improve the underlying decision process, making them particularly difficult for multimodal SLM to learn from. To address these issues, we propose Input Attribution-Aware Policy Optimization (IAPO), an RL algorithm for improving tool use in multimodal SLM by aligning the model's attribution across input components with that of a stronger teacher. Experiments on Qwen2.5-VL-3B show that the proposed method improves visual question answering accuracy by an average of 3% across six test sets compared with existing visual tool use work, by helping the model attend to the most relevant input evidence.

阅读与讨论 → 访问原文 →

03.

arXiv (CS.AI) 2026-06-19 DOI: arXiv:2606.19386

Bistable by Construction: Wall-Clock-Calibrated State Monitors Have No Moment-Detection Regime at Agent Cadence

作者:

Manvendra Modgil↗

arXiv:2606.19386v1 Announce Type: cross Abstract: Runtime monitors for autonomous agents commonly threshold an accumulated internal state - a behavioural baseline, a drift statistic, or, in our prior work, a modelled affective state. We previously reported a State Saturation Trap: threshold-on-state triggers over a continuous affect engine become near-constant alarms on SWE-bench debugging agents (Modgil 2026). A post-release audit found the engine received dt=0 between actions, so its exponential decay never operated: the published trap is a pure-accumulator result. We correct the record (erratum, v2) and treat the flaw as an experiment. The key variable it exposes is whether a monitor's dynamics are calibrated in sample time (per observation, as in CUSUM) or wall-clock time (half-lives in seconds, as in affect models and EMA baselines). On fixed-rate streams these coincide; on agent streams, where inter-action time varies by orders of magnitude, they do not. A pre-registered sweep over uniform intervals (dt in {0..600}s) on 20 trajectories shows the wall-clock level trigger has two regimes: at dt=60s silent. Every critical dt lies in (1,30]s. Real agent runs measure latency at median 1.53s (p90 2.33s); real coding cadence sits inside the trap regime, vindicating the empirical finding under a corrected mechanism. The structure is a property of the calibration class, not the engine: a minimal wall-clock accumulator over the raw error stream reproduces the same cliff, while a sample-time CUSUM over the identical stream is exactly dt-invariant (20/20). A rising-edge trigger with hysteresis fires 0-3 times per trajectory in every condition. We conclude that wall-clock-calibrated leaky-integrator monitors admit no regime in which they act as moment detectors on agent streams; transition detection escapes the trap at every cadence, but does not recover human intervention timing.

阅读与讨论 → 访问原文 →

04.

arXiv (CS.CL) 2026-06-11 DOI: arXiv:2606.10820

K-Forcing: Joint Next-K-Token Decoding via Push-Forward Language Modeling

作者:

Zhiwei Tang↗Yuanyu He↗Yizheng Han↗Wangbo Zhao↗Jiasheng Tang↗Fan Wang↗Bohan Zhuang↗

Autoregressive (AR) language modeling is the dominant paradigm for text generation, yet its sequential token-by-token decoding makes inference memory-bound and inefficient. Existing acceleration approaches, such as speculative decoding and diffusion language models, can yield speedups under certain conditions but do not directly address high-load batch serving–the scenario most critical for industrial-scale deployment. We introduce K-Forcing, a push-forward language modeling paradigm for joint next-k-token decoding. K-Forcing distills an existing AR model into a conditional push-forward mapping–one that transforms independent uniform noise variables into a joint sample of multiple future tokens in a single forward pass. This design preserves fixed-length outputs, reuses the AR teacher backbone, and remains compatible with standard AR serving infrastructure. We train this mapping via progressive self-forcing distillation, which gradually expands the prediction window while enabling the student to closely match the sequence distribution of the AR teacher. We evaluate K-Forcing on LM1B and OpenWebText using a standard causal Transformer backbone. When aggressively configured to generate k = 4 tokens per forward pass, K-Forcing delivers approximately 2.4-3.5x speedup across different batch sizes, while incurring modest quality degradation relative to its AR teacher. As inference increasingly dominates the lifetime compute cost of modern LLMs, K-Forcing offers a promising route toward accelerating AR generation under real-world high-load deployment.

阅读与讨论 → 访问原文 →

05.

arXiv (CS.AI) 2026-06-19 DOI: arXiv:2606.20236

A Multi-Agent system for Multi-Objective constrained optimization

作者:

Federica Filippini↗

arXiv:2606.20236v1 Announce Type: new Abstract: Many decision-making problems in computing and networking systems can be naturally formulated as cost-minimization problems under performance constraints. In dynamic environments, reinforcement learning (RL) is often used to solve such problems at runtime by embedding both costs and constraint violations into a single scalar reward through weighted penalty terms, following a Lagrangian-inspired formulation. However, in this context the behavior of the learned policy critically depends on the choice of these weights, which are typically selected manually. This makes it difficult to identify an appropriate trade-off between optimizing the primary objective and effectively avoiding constraint violations, particularly in non-stationary environments where their relative importance may change. This paper presents MAMO (Multi-Agent system for Multi-Objective constrained optimization), an approach to tackle this balancing problem through multi-agent RL. MAMO decouples task execution from objective design by formulating the selection of reward weights as a learning problem, providing a !rst step towards more autonomous and robust RL-based solutions for constrained optimization problems in dynamic environments.

阅读与讨论 → 访问原文 →

06.

arXiv (CS.LG) 2026-06-16 DOI: arXiv:2606.15569

A Decision-Theoretic View of Test-Time Training: When, How Far, and Which Directions to Adapt

作者:

Tomoya Wakayama↗

arXiv:2606.15569v1 Announce Type: new Abstract: Test-time training (TTT) adapts a pretrained model to each prompt via parameter updates, improving accuracy under pretraining-to-test distribution shifts. Yet, its performance often suffers from instability and sensitivity to hyperparameters such as update steps and subspace. We explain this behavior through a decision-theoretic lens, treating TTT as implicit Bayesian inference in the kernel regime. Under a Gaussian process benchmark, we show that TTT reduces prediction error when updates are spectrally matched to the prompt's signal-to-noise ratio and aligned with query-relevant eigen-directions. This perspective underpins the following results: (1) we show when fixed update steps and subspaces fail under distribution shifts, motivating adaptive strategies; (2) we prove that selecting update steps via prompt evidence admits a PAC-Bayes guarantee against overfitting; and (3) we characterize the Bayes-optimal update subspace under a linear-Gaussian correction model, yielding a scoring rule for selecting Transformer blocks and heads. Our theory helps explain the empirical instability of TTT, taking a step toward principled guidance for when, how far, and which directions to adapt.

阅读与讨论 → 访问原文 →

07.

arXiv (CS.AI) 2026-06-15 DOI: arXiv:2501.08561

ANSR-DT: A Neuro-Symbolic Framework for Adaptive and Explainable Digital Twins

作者:

Safayat Bin Hakim↗Muhammad Adil↗Alvaro Velasquez↗Houbing Herbert Song↗

arXiv:2501.08561v4 Announce Type: replace Abstract: Digital twins are increasingly used to monitor and optimize industrial systems, yet many existing frameworks remain difficult to interpret, slow to adapt, and limited in their ability to incorporate explicit domain knowledge. This paper presents ANSR-DT, an adaptive neuro-symbolic framework that unifies temporal anomaly detection, symbolic reasoning, and reinforcement-learning-based decision support within a single digital twin pipeline. ANSR-DT combines a CNN-LSTM model for multivariate pattern recognition with Prolog-based reasoning that converts learned signals into explicit rules, enabling transparent diagnoses and traceable decision paths. A PPO-based adaptation layer further refines operational responses under changing conditions while preserving interpretability. Experiments against 8 baselines show that ANSR-DT delivers competitive predictive performance together with stable rule extraction, scalable symbolic reasoning, and actionable explanations. Additional validation on the Skoltech Anomaly Benchmark (SKAB) further indicates that the framework transfers beyond synthetic settings. These findings position ANSR-DT as a practical foundation for trustworthy, adaptive, and explainable industrial digital twins.

阅读与讨论 → 访问原文 →

08.

arXiv (CS.CL) 2026-06-12 DOI: arXiv:2605.22641

More Context, Larger Models, or Moral Knowledge? A Systematic Study of Schwartz Value Detection in Political Texts

作者:

V\'ictor Yeste↗Paolo Rosso↗

Detecting Schwartz values in political text is difficult because implicit cues often depend on surrounding arguments and fine-grained distinctions between neighboring values. We study when context and explicit moral knowledge help sentence-level value detection. Using the ValuesML/Touché ValueEval format, we compare sentence, window, and full-document inputs; no-RAG and retrieval-augmented settings with a curated moral knowledge base; supervised DeBERTa-v3-base/large encoders; and zero-shot LLMs from 12B to 123B parameters. The results show that more context is not uniformly better: full-document context improves supervised DeBERTa encoders by 3.8-4.8 macro-F1 points over sentence-only input, but does not consistently help zero-shot LLMs. Retrieved moral knowledge is more consistently useful in matched comparisons, improving each tested model family and context condition under early fusion. However, scaling from DeBERTa-v3-base to large and from 12B to larger LLMs does not guarantee gains, and simple early fusion outperforms the tested late-fusion and cross-attention RAG variants for encoders. Per-value analyses show that context and retrieval help most for socially situated or conceptually confusable values. These findings suggest that value-sensitive NLP should evaluate context, knowledge, and model family jointly rather than treating longer inputs or larger models as universal improvements.

阅读与讨论 → 访问原文 →

09.

arXiv (CS.AI) 2026-06-19 DOI: arXiv:2603.28387

The Scaffold Effect: How Prompt Framing Drives Apparent Multimodal Gains in Clinical VLM Evaluation

作者:

Doan Nam Long Vu↗Simone Balloccu↗

arXiv:2603.28387v2 Announce Type: replace Abstract: Trustworthy clinical AI requires that performance gains reflect genuine evidence integration rather than surface-level artifacts. We evaluate 12 open-weight vision-language models (VLMs) on binary classification across two clinical neuroimaging cohorts, \textsc{FOR2107} (affective disorders) and \textsc{OASIS-3} (cognitive decline). Both datasets come with structural MRI data that carries no reliable individual-level diagnostic signal. Under these conditions, smaller VLMs exhibit gains of up to 58\% F1 upon introduction of neuroimaging context, with distilled models becoming competitive with counterparts an order of magnitude larger. A contrastive confidence analysis reveals that merely mentioning MRI availability in the task prompt accounts for 70-80\% of this shift, independent of whether imaging data is present, a domain-specific instance of modality collapse we term the scaffold effect. Expert evaluation reveals fabrication of neuroimaging-grounded justifications across all conditions, and preference alignment, while eliminating MRI-referencing behavior, collapses both conditions toward random baseline. Our findings demonstrate that surface evaluations are inadequate indicators of multimodal reasoning, with direct implications for the deployment of VLMs in clinical settings.

阅读与讨论 → 访问原文 →

10.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2606.18363

Guava: An Effective and Universal Harness for Embodied Manipulation

作者:

Haowen Liu↗Xirui Li↗Shaoxiong Yao↗Peng Shi↗Tianyi Zhou↗Jia-Bin Huang↗Furong Huang↗Jiayuan Mao↗

arXiv:2606.18363v1 Announce Type: cross Abstract: Language models trained on large-scale vision-language data have demonstrated strong potential for embodied agents. Harnessing models through embodied tools use offers a promising alternative to end-to-end vision-language-action systems by combining high-level reasoning with external modules for perception, planning, and control. However, it remains unclear what makes an effective harness for embodied manipulation, and to what extent such a harness can unlock embodied capabilities in a wide range of reasoning models. In this work, we present Guava, a harness framework for embodied tool use developed through systematic exploration of the design space of agent workflows, action spaces, and observation spaces. Our study identifies three key ingredients for effective embodied agents: iterative perception-reasoning-action loops, semantic action abstractions, and multimodal observations. To understand whether these design principles are universal even to small models, we develop an end-to-end training pipeline that distills embodied manipulation capabilities into a 4B open-source model using fewer than 2K trajectories collected entirely in simulation. Experimental results in both simulation and real-world environments show performance comparable to frontier proprietary models while exhibiting strong generalization to unseen objects, novel instructions, and long-horizon tasks. Results suggest that a well-designed harness can serve as a scalable, model-agnostic interface for embodied manipulation, enabling strong emergent embodied capabilities in compact open-source models with minimal training data.

阅读与讨论 → 访问原文 →

11.

arXiv (CS.CV) 2026-06-16 DOI: arXiv:2606.14765

Momentum-Guided Semantic Forecasting (MoFore) for Self-Supervised Video Representation Learning

作者:

Qinwu Xu↗

Self-supervised video representation learning has recently advanced through contrastive learning, masked reconstruction, and predictive representation learning. Reconstruction-based approaches such as MAE and VideoMAE learn representations by recovering masked visual content [he2022mae,tong2022videomae], while contrastive methods such as CLIP learn semantically meaningful embedding spaces through representation alignment [radford2021clip]. In this work, we introduce a Momentum-Guided Semantic Forecasting framework (MoFore) for self-supervised video representation learning. Instead of optimizing for pixel-level reconstruction or task-specific semantic alignment, the proposed method learns temporally predictive video representations by forecasting future latent embeddings from temporally distant context clips. To improve robustness across temporal scales, we further introduce randomized temporal-gap forecasting during training. The framework combines predictive latent forecasting with contrastive regularization to encourage temporal consistency while preventing representation collapse. Experiments on the UCF101 dataset demonstrate that the proposed framework learns temporally consistent and semantically meaningful video representations without using action labels during training. Quantitative analysis shows strong temporal stability and emergent category-level structure in the learned embedding space, while qualitative retrieval experiments reveal motion-aware organization across related activities. Overall, the results suggest that long-range latent forecasting provides an effective and computationally efficient approach for self-supervised video representation learning without relying on reconstruction-based objectives.

阅读与讨论 → 访问原文 →

12.

arXiv (CS.LG) 2026-06-19 DOI: arXiv:2606.20546

Predictability as a Fine-Grained Measure for Privacy

作者:

Linda Lu↗Karthik Sridharan↗

arXiv:2606.20546v1 Announce Type: new Abstract: Differential privacy (DP) ensures rigorous individual-level privacy guarantees against even the most knowledgeable attackers, but its worst-case nature can impose a costly privacy-accuracy tradeoff. We introduce privacy via predictability, a fine-grained framework that explicitly incorporates the attacker's core knowledge, a compromised portion of the dataset generated by a stochastic process, and a specified family of queries. Predictability measures privacy leakage as the incremental gain in an attacker's ability to predict sensitive information about unknown individuals after observing the algorithm's output, beyond what can already be inferred from the compromised data. We show that predictability and DP are generally incomparable: each can be small while the other is large. However, in the worst-case regime where all but one individual is compromised, and all binary queries are considered sensitive, predictability implies mutual-information DP. More generally, predictability provides a finer-grained privacy metric tailored to specific sensitive information and specific attacker models. We introduce a general framework, using the generalized method of moments (GMM), to analyze asymptotic predictability when the compromised data is generated by a stationary, ergodic, mixing process. Using this analysis, we derive a predictability-calibrated output perturbation scheme for ERM. Our approach is complementary to DP and can be used alongside DP to provide fine-grained privacy control.

阅读与讨论 → 访问原文 →

13.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2605.26595

Cordyceps: Covert Control Attacks on LLMs via Data Poisoning

作者:

Zedian Shao↗Charles Fleming↗Teodora Baluta↗

arXiv:2605.26595v2 Announce Type: replace-cross Abstract: Large language models (LLMs) are often fine-tuned on uncurated text datasets that adversaries can poison. Existing poisoning attacks primarily rely on fixed trigger phrases that defenses such as outlier detection, clean-data regularization, or online monitoring can neutralize. In this paper, we propose a data poisoning method that teaches an LLM an information hiding scheme reliably and stealthily through semantic associations between shared knowledge such as facts or concepts and attacker-chosen phrases. The induced hiding scheme can encode and decode arbitrary malicious instructions, thus revealing a new and subtle poisoning-induced vulnerability: covert control attacks. We precisely characterize covert control attacks and evaluate them across $5$ LLMs, $3$ backdoor defenses, and $4$ prompt injection defenses. With a small poisoned fraction, covert control attacks outperform heuristic-based prompt injection attacks in average attack success rate by about $40\%$ relative to clean fine-tuned models. They also circumvent defenses based on detection and fine-tuning, maintaining up to $93\%$ attack success rate after backdoor defenses and up to $98\%$ after prompt injection defenses.

阅读与讨论 → 访问原文 →

14.

arXiv (CS.CL) 2026-06-18 DOI: arXiv:2606.18372

Redact or Keep? A Fully Local AI Cascade for Educational Dialogue De-Identification

作者:

Haocheng Zhang↗Zhuqian Zhou↗Kirk Vanacore↗Bakhtawar Ahtisham↗Ren\'e F. Kizilcec↗

Educational dialogue is a valuable but sensitive resource for research: the same transcripts that capture authentic learning often capture personally identifiable information (PII) entangled with curricular content, where "Riemann" may refer to a real student or to a mathematical concept. Existing approaches force a tradeoff between governance and accuracy. Commercial Large Language Models (LLMs) can handle this ambiguity but require sending student data to third parties, while local named entity recognition (NER) systems preserve governance but over-redact curricular terms. We propose a fully local cascade framework that reframes de-identification from open-ended entity recognition to constrained privacy triage. A recall-first union proposer combines two lightweight encoders with deterministic rules to over-generate candidate spans; a context-aware reviewer then makes a binary Redact/Keep decision for each candidate using surrounding dialogue and speaker role. We evaluate three reviewer configurations against same-family LLM-only baselines and a commercial API on math tutoring transcripts from two large platforms. The strongest local configuration reaches 0.958 macro F1, compared with 0.767 for a same-family LLM-only baseline and 0.706 for the commercial API, while running entirely on a single laptop. On a targeted challenge set of curricular-personal name ambiguity, the same configuration degrades by only 0.03 F1 versus 0.19 to 0.25 for smaller reviewers. These results suggest that for educational de-identification, problem formulation matters more than model scale.

阅读与讨论 → 访问原文 →

15.

arXiv (CS.CL) 2026-06-11 DOI: arXiv:2606.12370

Breaking Entropy Bounds: Accelerating RL Training via MTP with Rejection Sampling

作者:

Yucheng Li↗Huiqiang Jiang↗Yang Xu↗Jianxin Yang↗Yi Zhang↗Yizhong Cao↗Yuhao Shen↗Fan Zhou↗Rui Men↗Jianwei Zhang↗An Yang↗Bowen Yu↗…

Reinforcement learning (RL) has become a key component in modern large language models, yet the rollout stage remains the key bottleneck in RL training pipelines. Although Multi-Token Prediction (MTP) offers a natural solution to accelerate rollouts through speculative decoding, many studies have observed that MTP acceptance rates degrade significantly during RL training, leading to limited speedup performance. To address this bottleneck, we present Bebop, a systematic study of MTP in LLM post-training, and offer practical recipes to integrate MTP into large-scale RL pipelines. First, we reveal that the MTP acceptance rate is fundamentally bounded by the fluctuation of model entropy, which demonstrates a clear negative linear relationship with the rise of entropy in the RL stage. Second, we show that probabilistic rejection sampling largely alleviates the disturbance introduced by entropy in RL compared to greedy draft sampling. We further identify that the conventional MTP training objectives (cross-entropy or KL) are suboptimal in such settings, and therefore we propose a novel end-to-end TV loss that directly optimizes multi-step rejection sampling acceptance rate, yielding ~10% acceptance rate improvements, achieving up to 95% acceptance rates and up to 25% extra inference throughput gains across mathematical reasoning, code generation, and agentic tasks. Third, we test various online MTP training strategies during RL and show that pre-RL MTP training with e2e TV loss and rejection sampling achieves a consistent acceptance rate and speedup throughout the entire RL, eliminating the need for costly online MTP updating. We provide extensive experiments and analysis that validate our findings. Experimental results show our method achieves up to 1.8x end-to-end acceleration in async RL training of Qwen3.5, Qwen3.6, and Qwen3.7 models.

阅读与讨论 → 访问原文 →

16.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2606.16560

The BD-LSC Dataset: Facilitating the Benchmarking of Models for Lexical Semantic Change Detection in Slang and Standard Usage

作者:

Afnan Aloraini↗Viktor Schlegel↗Goran Nenadic↗Riza Batista-Navarro↗

Automatic semantic change detection aims to identify how word meanings shift over time, offering insights into both linguistic and societal change. Despite recent progress in computational lexical semantic change (LSC), existing benchmarks and methods struggle to capture bi-directional semantic change, particularly cases where words simultaneously gain and lose senses. This problem is especially challenging for words that have both slang and standard meanings. To address these gaps, we introduce two complementary benchmark datasets. The Bi-Directional Lexical Semantic Change (BD-LSC) dataset captures sense gain, sense loss, and stability across three time periods, enabling the study of complex semantic trajectories. The SlangTrack Word Sense Disambiguation (ST-WSD) dataset provides fine-grained, instance-level sense annotations for words combining slang and standard usages, supporting systematic benchmarking of WSD and semantic change detection models. Using these benchmarks, we systematically evaluate models across different methodological families: unsupervised clustering using contextualised embeddings, supervised machine learning, transformer-based models, and state-of-the-art large language models. Among the evaluated systems, the few-shot GPT-4o model achieved the strongest aggregate performance on Exact Sense Match (ESM) and multi-label accuracy; however, Macro-F1 scores near 0.5 across all systems show that rare slang senses remain difficult, which we identify as the central open challenge.

阅读与讨论 → 访问原文 →

17.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2605.09169

Prediction Bottlenecks Don't Discover Causal Structure (But Here's What They Actually Do)

作者:

Ankit Hemant Lade↗Sai Krishna Jasti↗Indar Kumar↗Aman Chadha↗

arXiv:2605.09169v2 Announce Type: replace-cross Abstract: A Mamba state-space model trained only for next-step prediction appears to recover Granger-causal structure through a simple readout $S = |W_{out} W_{in}|$, with early experiments suggesting the phenomenon generalized across architectures and benefited from interventional data at $p < 10^{-5}$. We package the protocol used to test that claim – standardized synthetic generators (VAR/Lorenz/CauseMe-style), three intervention semantics ($do(X=c)$, soft-noise, random-forcing), edge-provenance cards on three real datasets, and size-matched control arms – as a reusable falsification benchmark, and walk the claim through it in five stages. The method-level claim does not survive: (i) a plain linear bottleneck does as well or better; (ii) tuned Lasso beats the bottleneck on synthetic CauseMe-style benchmarks, and on Lorenz-96 (the only real benchmark with unambiguous ground truth) classical PCMCI and Granger lead a tight cluster in which the bottleneck trails; (iii) the headline intervention advantage is roughly 60% a sample-size confound, and the residual disappears under standard $do(X=c)$ interventions, surviving only under a non-standard random-forcing scheme; (iv) even that residual reproduces, with a larger effect, in classical bivariate Granger – the effect is method-agnostic. What survives is a narrow characterization result; the benchmark is the lasting artifact, and each stage above is one of its control arms.

阅读与讨论 → 访问原文 →

18.

arXiv (CS.AI) 2026-06-19 DOI: arXiv:2606.19950

Confidence Calibration for Multimodal LLMs: An Empirical Study through Medical VQA

作者:

Yuetian Du↗Yucheng Wang↗Ming Kong↗Tian Liang↗Qiang Long↗Bingdi Chen↗Qiang Zhu↗

arXiv:2606.19950v1 Announce Type: cross Abstract: Multimodal Large Language Models (MLLMs) show great potential in medical tasks, but their elicited confidence often misaligns with actual accuracy, potentially leading to misdiagnosis or overlooking correct advice. This study presents the first comprehensive analysis of the relationship between accuracy and confidence in medical MLLMs. It proposes a novel method that combines Multi-Strategy Fusion-Based Interrogation (MS-FBI) with auxiliary expert LLM assessment, aiming to improve confidence calibration in Medical Visual Question Answering (VQA). Experiments demonstrate that our method reduces the Expected Calibration Error (ECE) by an average of 40\% across three Medical VQA datasets, significantly enhancing MLLMs' reliability. The findings highlight the importance of domain-specific calibration for MLLMs in healthcare, offering a more trustworthy solution for AI-assisted diagnosis.

阅读与讨论 → 访问原文 →

19.

arXiv (CS.AI) 2026-06-19 DOI: arXiv:2606.20122

ScaffoldAgent: Utility-Guided Dynamic Outline Optimization for Open-Ended Deep Research

作者:

Zhibang Yang↗Xinke Jiang↗Yuzhen Xiao↗Ruizhe Zhang↗Yue Fang↗XinFei Wan↗Zhengxing Song↗Yuxuan Liu↗Yuheng Huang↗Xu Chu↗Junfeng Zhao↗Yasha Wang↗…

arXiv:2606.20122v1 Announce Type: new Abstract: Open-ended deep research (OEDR) requires systems to acquire knowledge through multi-round retrieval and generate coherent long-form reports. The outline plays a central role as a structural scaffold that coordinates retrieval, evidence organization, and generation. However, existing methods either fix the outline before writing or refine it with local heuristics, leading to scaffold drift under continuous information accumulation and delayed feedback for evaluating outline modifications. We propose ScaffoldAgent, a utility-guided dynamic outline optimization framework for OEDR. ScaffoldAgent models outline evolution as a structured decision process with three operations: Expansion, Contraction, and Revision, enabling controlled updates to the report scaffold. It further introduces a utility-guided feedback mechanism that estimates the downstream value of each outline operation from retrieval gain, structural coherence, and trial-generation quality. The resulting utility signal guides node selection, operation scheduling, and termination during inference. Experiments on DeepResearch Bench and DeepResearch Gym show that ScaffoldAgent consistently improves long-form report generation and factual grounding over existing deep research agents.

阅读与讨论 → 访问原文 →

20.

arXiv (CS.LG) 2026-06-18 DOI: arXiv:2606.19150

Complementary Attention Head Pruning for Efficient Transformers

作者:

Yaniv Livertovsky↗Shahar Somin↗Gonen Singer↗

arXiv:2606.19150v1 Announce Type: new Abstract: The remarkable success of Transformer-based models in natural language processing stems from architectural scaling, which leads to a large number of parameters and hinders deployment in resource-constrained environments. While structured pruning offers a pathway to compression, existing state-of-the-art methods often rely on gradient-based importance ranking or stochastic gating, which suffer from instability, structural degeneration, and the need for extensive manual hyperparameter tuning. In this paper, we introduce CAHP (Complementary Attention Head Pruning), a novel post-hoc framework that redefines head selection as a global graph-theoretical problem. Rather than evaluating heads in isolation, CAHP utilizes graph-based clustering combined with information-theoretic distance measures to identify and preserve a topologically diverse subset of complementary attention heads. Without requiring a predefined sparsity level or pruning ratio, the framework automatically determines the number of selected attention heads across layers by identifying a diminishing marginal performance curve, where pruning additional heads leads to a sharp degradation in performance, as determined by the chosen polynomial degree. Extensive evaluations on the SST-5 and MNLI benchmarks, across different Transformer model scales, demonstrate that CAHP consistently outperforms competitive baselines, particularly in high-compression regimes. Furthermore, our structural analysis shows that CAHP avoids the "proximity bias" of gradient-based pruning methods, which tend to preserve heads mainly in layers close to the output, and instead retains a functionally critical set of attention heads in the model's intermediate layers.

阅读与讨论 → 访问原文 →

21.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2510.06198

The Answer Lies Within: Self-Derived Rewards Enable Explainable Relation Extraction

作者:

Xinyu Guo↗Zhengliang Shi↗Minglai Yang↗Mihai Surdeanu↗

Despite the remarkable reasoning capabilities of large language models, they still struggle with one-shot relation extraction without predefined relation labels. We identify two pitfalls: models are often misled by irrelevant tokens instead of relation-conveying semantics, and they often fail to align with the abstraction level human annotators expect. We introduce a novel framework that closes this gap with two components: (1) COGRE, a cognitively-inspired reasoning framework that structures RE into a series of processes mimicking human text-processing; and (2) HIT@DICT, a reinforcement learning intermediate reward strategy that encourages reasoning to align with relational labels by rewarding relation-relevant phrases in reasoning. The reward is derived on a credit dictionary automatically extracted from correct predictions. Our experiments show that our framework improves both accuracy and explanation quality by addressing these two pitfalls. For example, COGRE with Qwen2.5-14B-Instruct on One-shot NYT29 achieves 24.65% F1, surpassing prior reasoning-based designs. Optimizing this approach with RL using HIT@DICT further improves performance by +23.46% points. Finally, human evaluation shows that our best model generates relational phrases closely aligned with gold labels, increasing human explanation quality ratings by 54% (relative).

阅读与讨论 → 访问原文 →

22.

arXiv (CS.AI) 2026-06-17 DOI: arXiv:2606.18023

LoopCoder-v2: Only Loop Once for Efficient Test-Time Computation Scaling

作者:

Jian Yang↗Shawn Guo↗Wei Zhang↗Tianyu Zheng↗Yaxin Du↗Haau-Sing Li↗Jiajun Wu↗Yue Song↗Yan Xing↗Qingsong Cai↗Zelong Huang↗Chuan Hao↗…

arXiv:2606.18023v1 Announce Type: cross Abstract: Looped Transformers scale latent computation by repeatedly applying shared blocks, but sequential looping increases latency and KV-cache memory with the loop count. Parallel loop Transformers (PLT) alleviate this cost through cross-loop position offsets (CLP) and shared-KV gated sliding-window attention, making loop count a practical design choice. We therefore study PLT loop-count selection through a gain–cost view: an extra loop may refine representations, but CLP also introduces a positional mismatch at each loop boundary. We instantiate this study by training LoopCoder-v2, a family of 7B PLT coders with different loop counts, from scratch on 18T tokens, followed by matched instruction tuning and evaluation. Empirically, the two-loop variant delivers broad gains over the non-looped baseline across code generation, code reasoning, agentic software engineering, and tool-use benchmarks, improving SWE-bench Verified from 43.0 to 64.4 points and Multi-SWE from 14.0 to 31.0 points. In contrast, variants with three or more loops regress, revealing a strongly non-monotonic loop-count effect. Our diagnostics show that loop 2 provides the main productive refinement, while later loops yield diminishing, oscillatory updates and reduced representational diversity. Because the CLP-induced mismatch remains roughly fixed as refinement gains shrink, the offset cost increasingly dominates. This gain–cost trade-off explains PLT's saturation at two loops and provides diagnostics for loop-count selection.

阅读与讨论 → 访问原文 →

23.

arXiv (CS.CL) 2026-06-19 DOI: arXiv:2606.19819

CREDENCE: Claim Reduction for Decomposition & Enhanced Credibility – Semantic Metrics and Convergence Analysis

作者:

Phuong Huu Vu Tran↗Thuan Duc Mai↗Bach Xuan Le↗

Decomposing compound sentences into atomic, verifiable claims is a prerequisite for reliable automated fact-checking. Prior work has relied on token-overlap (Jaccard) metrics that systematically underestimate decomposition quality for paraphrastic claims, and has lacked formal termination analysis for the repair loop. We present Credence, a revised claim decomposition and evaluation framework addressing both shortcomings. Our contributions are: (1) Semantic-F1: we use BGE-large cosine similarity fidelity metric that resolves Jaccard's penalisation and improves downstream fact-checking accuracy; (2) Convergence theorems: we formally characterise four properties of the repair pipeline, establishing that rule-based repair is monotone and finitely terminating under an oracle parser assumption; LLM-based self-repair is provably non-monotone and requires an early-exit guard; (3) Three evaluation benchmarks spanning social-media, encyclopaedic, and news domains for cross-domain generalisation measurement; (4) Multi-model benchmarking across four decomposer models (3.8B-12B) and a closed API model. Experiments on SocialClaimSplit, WikiSplitBench, and ClaimDecompBench show that Semantic-F1 outperforms Jaccard-F1 by +15-32pp. EPR ranges from 0.94 to 1.00 on SocialClaimSplit and WikiSplitBench, while ClaimDecompBench includes lower base EPR cases (down to 0.824) due to harder news-domain constructions, and rule-repair reduces the Atomicity Violation Rate (AVR) by 47-100% relative to the base model without degrading fidelity.

阅读与讨论 → 访问原文 →

24.

arXiv (CS.LG) 2026-06-16 DOI: arXiv:2606.16044

Circuit Tracing in Autoregressive Protein Language Models

作者:

Darin Tsui↗William Deinzer↗Daniel Saeedi↗Amirali Aghazadeh↗

arXiv:2606.16044v1 Announce Type: new Abstract: Protein language models (pLMs) can generate novel protein sequences with properties beyond those observed in nature, yet the mechanisms underlying protein generation remain poorly understood. Existing mechanistic interpretability methods based on sparse autoencoders and transcoders primarily focus on protein representation learning models and do not capture the computation required for autoregressive generation. Here, we introduce ProGenMech, a mechanistic interpretability framework for generative protein language models that extends cross-layer transcoders (CLTs) to ProGen3, a sparse Mixture-of-Experts model trained for both causal generation and span infilling. Unlike per-layer approaches, CLTs reconstruct each layer using sparse latent variables from all preceding layers, enabling faithful recovery of inter-layer generative computation. We further develop a zero-shot circuit discovery framework to identify sparse latent circuits responsible for protein generation and fitness prediction. In causal generation and zero-shot fitness estimation tasks, ProGenMech outperforms local transcoder baselines in recovering ProGen3's probability distribution and functional scoring behavior, while matching the original model's generative distribution in span infilling tasks. Moreover, the recovered circuits reveal biologically meaningful motifs and functional regions associated with conserved sequence patterns and protein fitness landscapes, establishing a foundation for interpretable and steerable protein generation.

阅读与讨论 → 访问原文 →

25.

arXiv (CS.LG) 2026-06-16 DOI: arXiv:2601.22642

Pushing the Boundaries of Natural Reasoning: Interleaved Bonus from Formal-Logic Verification

作者:

Chuxue Cao↗Jinluan Yang↗Haoran Li↗Kunhao Pan↗Zijian Zhao↗Zhengyu Chen↗Yuchen Tian↗Lijun Wu↗Conghui He↗Sirui Han↗Yike Guo↗

arXiv:2601.22642v2 Announce Type: replace Abstract: Large Language Models (LLMs) show remarkable capabilities, yet their stochastic next-token prediction creates logical inconsistencies and reward hacking that formal symbolic systems avoid. To bridge this gap, we introduce a formal logic verification-guided framework that dynamically interleaves formal symbolic verification with the natural language generation process, providing real-time feedback to detect and rectify errors as they occur. Distinguished from previous neuro-symbolic methods limited by passive post-hoc validation, our approach actively penalizes intermediate fallacies during the reasoning chain. We operationalize this framework via a novel two-stage training pipeline that synergizes formal logic verification-guided supervised fine-tuning and policy optimization. Extensive evaluation on six benchmarks spanning mathematical, logical, and general reasoning demonstrates that our 7B and 14B models outperform state-of-the-art baselines by average margins of 10.4% and 14.2%, respectively. These results validate that formal verification can serve as a scalable mechanism to significantly push the performance boundaries of advanced LLM reasoning.

阅读与讨论 → 访问原文 →