Academic Intelligence · Curated Daily

探索全球前沿学术脉络

AcademicHub 汇聚顶级期刊与预印本平台的实时文献。定制您的专属科研雷达,利用大语言模型自动生成交叉领域文献分析简报。

01.
arXiv (CS.AI) 2026-06-12

What Type of Inference is Active Inference?

arXiv:2606.04935v2 Announce Type: replace Abstract: Active inference casts decision-making as inference, with the Expected Free Energy (EFE) unifying goal-directed and information-seeking behavior. Recent work showed that EFE minimization can be written as Variational Free Energy (VFE) minimization on a generative model augmented with epistemic priors. We prove that the VFE of the augmented model can be rewritten as the VFE of the predictive model plus explicit entropy-correction terms, making the EFE contribution transparent. We then show that proper EFE-based planning requires combining these epistemic corrections with a planning correction that turns marginal inference into policy optimization, yielding a full variational characterization of EFE-based planning. This clarifies which corrections are needed for cross-entropy planning and for full EFE-based planning. The same entropy-corrected formulation leads to a detailed message-passing scheme for EFE-based planning together with simpler ablations. Experiments on three grid-world environments show that full EFE-based planning outperforms ablations that omit either the planning correction or the epistemic corrections.

02.
arXiv (CS.CL) 2026-06-16

Beyond Monolingual Deep Research: Evaluating Agents and Retrievers with Cross-Lingual BrowseComp-Plus

Deep research agents are increasingly evaluated on their ability to search for evidence, reason over retrieved sources, and produce grounded answers. Existing browsing benchmarks, however, largely assume that the user's query and the supporting evidence are written in the same language, leaving open whether agentic search systems can operate when relevant evidence appears in another language. We introduce XBCP (Cross-lingual BrowseComp-Plus), a controlled benchmark that preserves the English question-and-answer space of BrowseComp-Plus but varies the languages of the supporting documents. XBCP instantiates two complementary settings: in the cross-lingual setting, each query is paired with evidence in a single assigned language. In the multilingual setting, the full evidence corpus is distributed equally and randomly across 12 languages spanning high-resource and low-resource regimes. We evaluate four deep research agents using sparse and dense multilingual retrievers, measuring answer accuracy, evidence recall, search behavior, calibration, citation fidelity, and oracle retrieval. Results reveal substantial degradation when evidence is translated. Even strong, dense retrievers lose evidence recall, and agents become less calibrated and cite evidence less reliably. Notably, accuracy remains lower even when all gold evidence is supplied directly. These findings suggest that cross-lingual deep research exposes both retrieval failures and an independent, agent-side difficulty in integrating language-mismatched evidence.

03.
arXiv (CS.CV) 2026-06-19

Composed Object Retrieval: Object-level Retrieval via Composed Expressions

Retrieving fine-grained visual content based on user intent remains a challenge in multimodal systems. Although current Composed Image Retrieval (CIR) methods combine reference images with retrieval texts, they are constrained to image-level matching and cannot localize specific objects. To this end, we propose Composed Object Retrieval (COR), a new object-level retrieval task that retrieves target object(s) from candidate objects in a target image and grounds the retrieved result with pixel-level masks. Given a reference object, its mask, a target image, and a retrieval text describing the desired modification, COR requires models to perform composed visual-textual reasoning rather than relying on explicit category names. This setting introduces several challenges, including fine-grained compositional matching, negative-object filtering under visually similar distractors, and flexible single- or multi-object retrieval. We construct COR125K, the first large-scale COR benchmark, containing 125,541 retrieval triplets across 408 categories with base/novel splits for evaluating category-level generalization. We also present CORE, a unified end-to-end model that integrates reference region encoding, adaptive vision-text interaction, and region-level contrastive learning to align composed representations with target objects while suppressing background and distractors. Extensive experiments demonstrate that CORE significantly outperforms existing CIR-based pipelines and strong baselines in both base and novel categories, establishing a simple and effective foundation for fine-grained object-level multimodal retrieval. Code will be released publicly at https://github.com/wangtong627/COR.

04.
arXiv (CS.AI) 2026-06-16

Relational Structural Causal Models

arXiv:2606.14892v1 Announce Type: new Abstract: An artificial intelligence must have a model of its environment that is causal, supporting reasoning about interventions and counterfactuals, and also combinatorial, supporting generalization to unseen combinations of objects. In this work, we formally study when and how such a model can be learned. We develop relational structural causal models, extending structural causal models (Pearl 2009) to settings where objects and their relations vary. First, we show how answers to not only causal but also observational queries about unseen combinations of objects can not be identified without further assumptions. To enable such identification–including in the presence of unobserved confounding–we define relational causal graphs and derive symbolic identification criteria. Finally, we propose relational neural causal models, a provably correct approach that outperforms non-relational baselines on simulated traffic scenes with varying cars, signals, and pedestrians.

05.
arXiv (CS.CV) 2026-06-15

SED:Lightweight Saliency prediction for Event-based data via Distillation

Event-based saliency prediction has gained attention recently, as combining event cameras with saliency estimation can act as an upstream stage that naturally improves the efficiency of downstream eventbased perception at the edge. However, current approaches are either neuromorphic, underperforming on event-based saliency benchmarks, or too heavy for resource-constrained edge applications due to their reliance on transformers or 3D convolutions. Drawing inspiration from efficient convolutional modules, SED and aiming to exploit the temporal information in event data, we propose a lightweight network, trained through knowledge distillation, built on a Depthwise Spatio-Temporal Block (DSTconv) – a factorization of the 3D depthwise separable convolution. Relative to its teacher, our model reduces the model size from 180 MB to 0.32 MB (562x) and the parameter count from 45M to 81k (554x), while matching or outperforming it on the N-DHF1K and N-UCF Sports datasets. Moreover, it generalizes strongly beyond its training distribution, transferring from synthetic to real event data where a model trained from scratch fails.

06.
arXiv (CS.AI) 2026-06-11

LSTM based IoT Device Identification

arXiv:2304.13905v2 Announce Type: replace-cross Abstract: While the use of the Internet of Things is becoming more and more popular, many security vulnerabilities are emerging with the large number of devices being introduced to the market. In this environment, IoT device identification methods provide a preventive security measure as an important factor in identifying these devices and detecting the vulnerabilities they suffer from. In this study, we present an end-to-end machine learning pipeline that identifies IoT devices in the Aalto university dataset (IoT devices captures) using Long Short-Term Memory (LSTM) networks. Raw network packet captures (PCAP) are processed into 25 engineered features, which are then arranged as sliding-window time-series sequences. We systematically evaluate sequence lengths from 2 to 20, reporting that performance improves approximately linearly up to length 6 and thereafter in a wave-like pattern, reaching its peak at length 18. On the final held-out test set with the optimal configuration, the model achieves an accuracy of 79.85% and a macro-averaged F1-score of 75.70% across 27 device classes.

07.
arXiv (CS.LG) 2026-06-16

How Much Capacity Does EEG Denoising Need? Ultra-Compact Networks reveal Benchmark Saturation and Metric-Utility Gap

arXiv:2606.08594v2 Announce Type: replace Abstract: Deep learning EEG denoising architectures have scaled from tens of thousands to tens of millions of parameters, yet no prior study has isolated model capacity as the experimental variable or tested whether reconstruction metrics predict downstream neural-signal utility. We address both gaps by fixing architecture, loss, data split, and training recipe while sweeping only channel width from 1.05K to 40.26K parameters in a minimal depthwise-separable convolutional U-Net. Models were evaluated on the EEGDenoiseNet benchmark, cross-dataset BCI transfer tests, controlled baseline retraining, and downstream motor-imagery classification with five decoder families across all nine BCI Competition IV-2a subjects. Reconstruction performance saturated by 3-6.5K parameters, with post-elbow gains of at most 0.015 correlation coefficient per log10-parameter unit. An 8.46M-parameter baseline retrained under the same pipeline matched the 40.26K compact variant on EOG–a 200x parameter gap yielding no advantage–while a Patch-Transformer control reproduced the same diminishing-return shape. Downstream evaluation exposed a classifier-dependent metric-utility gap: reconstruction-optimized denoising significantly degraded CSP+LDA classification across all nine subjects and three artifact types (best denoised accuracy 0.547 vs. 0.612 noisy baseline; Bonferroni p=0.0488), persisting on naturally recorded trials (Delta=-0.047; BH-FDR q=0.0049). End-to-end neural decoders showed variable or neutral effects. Standard EEG denoising benchmarks are saturated far below current model capacity, and reconstruction metrics do not predict BCI utility. Ultra-compact models at 33-46 KB and 1.27-2.61M FLOPs/segment are practical for edge deployment. These findings argue for capacity-controlled evaluation, harder task-aware benchmarks, and mandatory downstream validation.

08.
arXiv (CS.CV) 2026-06-17

Attention Alignment Between Humans and Vision-Language Models

Visual perception depends on top-down goals and bottom-up sensory mechanisms. Vision-language models implement both, allowing us to treat each component as a separable hypothesis about what drives where we look. We compared spatial attention maps from six vision-language models against human fixation heatmaps recorded on 200 images during two tasks (general description and social captioning). The six models spanned a 2$\times$2 factorial of CNN vs.\ ViT encoders crossed with LSTM vs.\ Transformer decoders, plus Molmo 7B-D and Qwen3.5 9B. We found that both decoder and encoder architecture shaped alignment, but decoder choice dominated. LSTM vs.\ Transformer decoders increased alignment by 40–50 percentage points (80–87\% vs.\ 40–59\% of the human noise ceiling). In contrast, CNN vs.\ ViT encoders contributed a secondary 5–20 point advantage depending on decoder family, with CNN-LSTM the most aligned model overall (85–87\%). Despite their alignment advantage, LSTM-decoder attention maps were spatially diffuse and minimally task-differentiated; ViT-Transformer, the weakest in alignment, showed the sharpest spatial concentration and strongest task differentiation. A hemispatial-neglect simulation confirmed that ablating attention impacted LSTM decoders more than Transformer decoders. In an exploratory extension using TRIBE-simulated synthetic neural responses, fixation alignment and neural relevance dissociate: CNN-Transformer attention maps better predicted synthetic brain activity despite lower fixation alignment, with attention maps best predicting early visual cortex. Together, top-down and bottom-up components trade off what they predict in behavioral and synthetic neural data.

09.
arXiv (CS.AI) 2026-06-17

Memory as a Wasting Asset: Pricing Flash Endurance for Embodied Agents, and the Limits of Doing So

arXiv:2606.18144v1 Announce Type: new Abstract: A robot's flash endurance is a non-renewable stock: every persisted write spends one of a few thousand program/erase cycles and never refills, yet no fielded robot memory system prices which memories are worth an erase cycle. We treat embodied memory as depreciating capital and price that stock with a single endurance shadow price $\eta$, which makes cost-minimizing placement across a RAM / on-board NVM / cloud hierarchy a threshold in a wear-augmented per-byte index. The index is cost-optimal whatever the sign of the value-write association $\chi$; only when $\chi > 0$ does the optimum turn non-monotone, sending a robot's most valuable memories off its flash. The pivot is thus empirical, and we measure $\chi$ on real robot logs at a pre-specified gate: its sign is a property of the deployment regime – positive on recurrent long-horizon manipulation ($\hat{\chi} \approx +1.0 \times 10^{-3}$, replicated at full power), null on a shorter-horizon suite, and negative on non-recurrent teleoperation. Two boundaries scope the result. The endurance budget is dormant on premium 3,000-P/E TLC at datasheet prices and binding on the commodity QLC/eMMC ($\sim$1,000 P/E) that cheaper edge robots run. And where it binds, a learned wear-aware controller only ties price-based routing on task value, because realized value is tier-invariant across RAM, NVM, and cloud: the rent governs device lifetime and cost, not task performance. Whether wear-aware placement improves task value remains open – $\chi$ is measured against a value proxy, and the non-monotone optimum, while proven, is not yet observed in data.

10.
arXiv (CS.AI) 2026-06-15

Can Editing 1 Neuron Fix Repetition Loops in LLMs?

arXiv:2606.13705v1 Announce Type: cross Abstract: Yes. Can it cure doom loops? Probably not. The Gemma 4 instruction-tuned models share a reproducible failure: on long factual enumeration prompts, such as listing every episode of a TV series, the 88 IAU constellations, or the 151 original Pokemon, they collapse into repetition, either a tight verbatim loop or a list whose entries decay onto a single answer. These loops occur at rates as high as 95% and survive prompt rewording, inference-engine changes, and most sampling adjustments. In this paper we explore whether this behavior is localized enough to remove by weight edits. To localize the cause, we use per-layer ablation and per-neuron attribution, then confirm the strongest candidates with full-generation sweeps. The loops trace to a small set of MLP neurons (or, in the 26B-A4B Mixture-of-Experts model, a few routed experts) which we suppress with static weight edits. These "surgeries" can be as small as a single sign-inverted neuron (in the E2B model). The size of the effective edits grows with model scale, but in all cases, the loop patterns can be addressed at normal generation budgets while preserving general-purpose benchmark scores. However, the edits do not solve everything: we also study longer thinking budgets, where the two larger models most visibly enter doom looping, i.e. a non-convergent regime in which the model self-corrects in circles over a fact it cannot recall, exhausting the budget without committing to a final answer. We show this residual failure is reduced but not eliminated by the same edits, and argue it is fundamentally a knowledge-precision problem rather than a removable circuit; weight surgery can delete a loop, but it cannot supply a missing fact. Our results are both a feasibility demonstration, that is, evidence that a concrete generation pathology can be localized to a few parameters and edited out, and a delineation of where that approach stops.

11.
arXiv (CS.CL) 2026-06-11

SOMA-SQL: Resolving Multi-Source Ambiguity in NL-to-SQL via Synthetic Log and Execution Probing

Natural language interfaces to databases aim to translate user questions into executable SQL, yet remain brittle in real-world settings where questions are underspecified and schemas are large and ambiguous. Ambiguity across user questions, database schemas, and model interpretations are central failure modes in NL2SQL, leading to misaligned intent, incorrect schema grounding, and erroneous SQL generation. Existing approaches rely on human clarification or treat ambiguity as a schema representation problem, but these do not scale nor resolve ambiguity autonomously. We propose SOMA-SQL to automatically resolve ambiguity via targeted synthetic query log and ambiguity-driven probing. SOMA-SQL constructs synthetic query log to ground schema interpretation and guide candidate SQL generation; it then executes targeted probing queries, driven by a structured ambiguity taxonomy and candidate disagreements, to produce disambiguation evidence for final SQL selection and repair. This active approach to ambiguity discovery and resolution generalizes across unseen schemas and query distributions without human-in-the-loop. Experiments on six public benchmarks demonstrate that SOMA-SQL improves execution accuracy by 13.0% on average over state-of-the-art baselines, with gains of up to 16.7% on ambiguous questions.

12.
bioRxiv (Bioinfo) 2026-06-18

Identification of environmental factors and growth stages in the prediction of fibre yield and fibre quality traits in rain-grown cotton

Context Understanding how and when environmental conditions influence overall crop performance is crucial for optimising the development of genotypes to a specific breeding target environment. We focused on economically important traits of Australian rain-grown cotton including fibre yield and quality traits, which have not been investigated comprehensively. The aim of the study was to identify relevant environmental factors, and the timing and extent of their impact on rain-grown cotton production. Methods We used a data driven approach to analyse the relationship between ten climate related environmental factors across various plant growth stages and eight fibre yield and quality traits, using a large-scale field dataset of 9,283 records collected over 23 years at 4 locations, with 53 unique year-location combinations. We applied eight complementary statistical models including stepwise, penalised and Bayesian linear regression, regression-tree based ensemble methods and deep learning frameworks to (1) select the most essential environmental covariates affecting rain-grown cotton production, and (2) evaluate the predictive performance of these models. Results The environmental impacts on rain-grown cotton production were trait and growth-stage specific. Number of rainy days and solar radiation were identified as the most influential environmental factors for fibre yield traits, vapour pressure deficit at maximum daily temperature was the most influential factor for majority of fibre quality traits. However, each analysed trait was influenced by multiple environmental factors across multiple growth stages (rather than a single factor or a single growth stage). These influential covariates explained a wide range of variation in the traits, accounting for 5.8% to 68.2%. Using the best-fit random forest model, our findings revealed non-linear relationships between key environmental covariates and the traits. Conclusions Environmental factors at different rain-grown cotton growth stages are key determinants for the performance of end-of-season fibre yield and fibre quality parameters. These findings highlight the need to account for environment conditions when developing cotton varieties optimised for rain-grown production systems. Potential strategies are proposed whereby these key environmental factors can be used to increase the rate of genetic gain in rain-grown cotton production systems. Implications The results of this study will be crucial for future genetic evaluations and analyses of genotype-by-environment interaction effects in rain-grown cotton, which must account for the influence of the environment on plant performance. Furthermore, these methods can be applied to other species to identify critical growth stages and environmental factors which most influence crop performance.

13.
medRxiv (Medicine) 2026-06-12

Metastatic Patterns and Treatment Characteristics of Triple-Negative Breast Cancer in Nigeria: A Retrospective Cohort Study

Background: Triple-negative breast cancer (TNBC) is an aggressive breast cancer subtype characterized by the absence of estrogen receptor, progesterone receptor, and human epidermal growth factor receptor 2 expression. It is associated with limited targeted treatment options, early relapse, and a high propensity for visceral metastasis. Data describing metastatic patterns and treatment characteristics of TNBC in Nigeria remain limited. Methods: This retrospective descriptive cohort study included 869 patients with TNBC managed at the Medserve-LUTH Cancer Center, Lagos University Teaching Hospital, Nigeria between June 2019 and June 2024. Demographic, clinicopathologic, metastatic, and treatment-related data were extracted from electronic medical records. Descriptive statistics were used to summarize patient characteristics, metastatic patterns, and treatment profiles. Associations between metastatic disease and selected clinicopathologic and treatment variables were explored using Pearsons chi-square test. Complete-case analysis was applied throughout. Results: The mean age at presentation was 52.09 {+/-} 12.26 years. Most patients were married (79.1%), postmenopausal (64.3%), and of Yoruba ethnicity (56.8%). Advanced disease predominated, with Stage III and Stage IV disease accounting for 42.9% and 35.6% of cases, respectively. Invasive ductal carcinoma was the most common histologic subtype (77.0%), while Grade II tumours constituted 51.3% of graded cases. Surgery was performed in 73.1% of patients, predominantly mastectomy (70.9% of surgical procedures). Chemotherapy was administered to 83.2% of patients, most commonly anthracycline-based regimens (41.8%), while radiotherapy was delivered to 63.5% of patients, with hypofractionated schedules of 42-43 Gy in 15-16 fractions accounting for 47.2% of radiotherapy courses. Metastatic disease was documented in 32.9% of evaluable patients. Lung metastasis was the most frequent site (62.5%), followed by bone (46.3%), regional lymph node invasion (38.5%), liver (23.0%), and brain (22.6%). Tumour grade and histologic subtype were not significantly associated with metastatic disease, whereas radiotherapy exposure demonstrated a significant association with metastatic status ({chi}{superscript 2} = 10.35, p = 0.001). Conclusion: TNBC in this Nigerian cohort was characterized by advanced-stage presentation, invasive ductal predominance, extensive use of multimodality treatment, and substantial visceral metastatic burden. Lung metastasis was the most common metastatic site. These findings provide contemporary real-world data on TNBC in Nigeria and highlight the continuing need for earlier diagnosis, timely referral, and sustained investment in comprehensive cancer care services.

14.
arXiv (quant-ph) 2026-06-12

Trading symmetry for Hilbert-space dimension in Bell-inequality violation

arXiv:2601.02893v3 Announce Type: replace Abstract: In quantum information, asymmetry, i.e., the lack of symmetry, is a resource allowing one to accomplish certain tasks that are otherwise impossible. Similarly, in a Bell test using any given Bell inequality, the maximum violation achievable using quantum strategies respecting or disregarding a certain symmetry can be different. In this work, we focus on the symmetry involved in the exchange of parties and explore when we have to trade this symmetry for a lower-dimensional quantum strategy in achieving the maximal violation of given Bell inequalities. For the family of symmetric Collins-Gisin-Linden-Massar-Popescu inequalities, we provide evidence showing that there is no such trade-off. However, for several other Bell inequalities with a small number of dichotomic measurement settings, we show that symmetric quantum strategies in the minimal Hilbert space dimension can only lead to a suboptimal Bell violation. In other words, there exist symmetric Bell inequalities that can only be maximally violated by asymmetric quantum strategies of minimal dimension. In contrast, one can also find examples of asymmetric Bell inequalities that are maximally violated by symmetric correlations. The implications of these findings on the geometry of the set of quantum correlations and the possibility of performing self-testing therefrom are briefly discussed.

15.
arXiv (CS.AI) 2026-06-15

Regional Climate Model Emulation with Diffusion Approaches: What is the Added Value of Generative Machine Learning?

arXiv:2606.14570v1 Announce Type: cross Abstract: Emulators provide a cost-effective alternative to regional climate models (RCMs) by capturing their dynamical downscaling function. They link large-scale predictors simulated by global climate models (GCMs) to RCM-simulated high-resolution fields of the target variable, here precipitation. Machine learning methods, typically deep learning, are cheaper than running RCMs in computation time and energy. Among them, generative models are appealing because they can simulate ensembles of local high-resolution fields consistent with the predictors. This ensemble, which we call the uncertainty envelope, remains to be properly assessed for added value. Here, we make three contributions. First, we introduce ParamDiffusion, a new two-stage diffusion-based framework, and compare it with a state-of-the-art diffusion approach. Second, we expand standard validation through a comprehensive framework aligned with climate-science needs, examining specific precipitation events, including extremes. Third, within this framework, we assess the added value of diffusion approaches relative to deterministic methods. We intercompare four deep-learning models: a deterministic model designed to capture the precipitation tail; a parametric probabilistic model based on it; a recently proposed diffusion approach; and ParamDiffusion, which couples the parametric model with a diffusion model. Our results show that diffusion-based approaches reproduce climatological precipitation statistics with high skill, including distributional tails and spatially compounded extremes, while generating spatially detailed fields. However, none of the assessed models consistently accounts for the most extreme RCM-simulated events within its uncertainty envelope. Diffusion models are therefore promising for probabilistic RCM emulation, but progress is still required before they can reliably represent high-impact precipitation extremes.

16.
arXiv (CS.AI) 2026-06-19

Uncertainty-Aware Reward Modeling for Stable RLHF

arXiv:2606.19818v1 Announce Type: cross Abstract: Reinforcement learning from human feedback (RLHF) aligns large language models by training reward models on preference data and optimizing policies to maximize predicted rewards. However, this pipeline faces two fundamental challenges: (1) reward models cannot signal when their predictions are unreliable, since they usually act as deterministic point estimators; and (2) modern group-based policy optimization can amplify unreliable reward signals, as exemplified by GRPO's uniform treatment of rewards during advantage computation. As policies explore increasingly diverse responses, these two limitations create a critical vulnerability: unreliable reward estimates may be granted disproportionate influence, triggering severe reward hacking. We propose Uncertainty-Aware Reward Modeling (UARM), which equips reward models with calibrated uncertainty via quantile-based conformal prediction and reweights GRPO advantages through heteroscedastic variance decomposition. Experiments across HelpSteer, UltraFeedback, and PKU-SafeRLHF demonstrate that UARM significantly improves reward model calibration, reduces reward hacking, and enhances downstream alignment quality compared to standard GRPO and uncertainty-agnostic baselines.

17.
PLOS Computational Biology 2026-06-11

MicroRNA target gene prediction model based on input-feature dependency and sample data expansion technique

作者:

by Yan Shao, Yazhou Li, Hexin Zhai, Shimin Dong Predicting microRNA target genes is essential for understanding their biological functions. This study developed a miRNA target gene prediction model based on input-feature dependency. Features were treated as multiple random variables, with marginal densities estimated using Gaussian mixture models (GMM) and dependencies captured by regular vine (R-vine) copula to derive joint probability density functions. We constructed class-conditional joint densities for positive and negative samples separately using GMM and R-vine copula, then combined these with prior probabilities using Bayes’ rule to obtain posterior probabilities of positive interactions, using a standard 0.5 probability threshold for deterministic prediction. To address insufficient data and class imbalance, hybrid distribution mega-trend diffusion was used to generate virtual samples for data augmentation. Computational validation showed high predictive performance even when only 30% of the training data were used. As proof-of-concept, we experimentally validated one predicted interaction (miR-8485 targeting JAK2) using dual-luciferase, cellular, and animal experiments, confirming the biological relevance of this specific model-generated prediction. These findings provide a valuable tool for understanding miRNA functions and disease mechanisms.

18.
arXiv (CS.LG) 2026-06-16

Asymptotically Optimal Sequential Testing with Markovian Data

arXiv:2602.17587v3 Announce Type: replace-cross Abstract: We study one-sided and $\alpha$-correct sequential hypothesis testing for data generated by an ergodic, finite-state Markov chain. The null hypothesis is that the unknown transition matrix belongs to a prescribed set $P$ of stochastic matrices, and the alternative corresponds to a disjoint set $Q$. We establish a non-asymptotic instance-dependent lower bound on the expected stopping time of any valid sequential test under the alternative, which is asymptotically tight. Our novel analysis improves the existing lower bounds, which are either asymptotic or provably sub-optimal in this setting. Our lower bound incorporates both the stationary distribution and the transition structure induced by the unknown Markov chain. We further propose an optimal test whose expected stopping time matches this lower bound asymptotically as $\alpha \to 0$. We illustrate the usefulness of our framework through applications to sequential detection of model misspecification in Markov Chain Monte Carlo and to testing structural properties, such as the linearity of transition dynamics, in Markov decision processes. Our findings yield a sharp and general characterization of optimal sequential testing procedures under Markovian dependence.

19.
arXiv (CS.CL) 2026-06-17

A Framework for Evaluating Agentic Skills at Scale

Agent skills – structured, reusable knowledge artifacts that augment LLM agent capabilities – have been rapidly adopted in industry, yet their cross-domain impact and use across commercial and open-source models remain under-studied, and no reusable methodology exists for evaluating an individual skill. In this work, we present an evaluation framework that lets a skill author construct realistic tasks to rigorously assess the aspects of a skill that matter most to them, and that estimates skill utility by solving those tasks. Further, we apply our evaluation approach at scale to 500 real-world skills, generating 1,000 tasks derived from the skills' content, along with instruction-following and goal-completion scoring rubrics. Using these metrics, we evaluate how 19 agent-model configurations, both proprietary and open-source, perform on the tasks. Our results show that models vary widely in how closely they adhere to the instructions encoded in skills, leading to substantial differences in their performance gains. Furthermore, we show that access to a skill significantly changes model behavior compared to the no-skill setup, providing an essential mechanism for encoding opinionated workflows into LLM agents. We release our evaluation dataset to support future work on agent skills.

20.
arXiv (CS.CL) 2026-06-15

EiCAP: Beyond Fluency, Probing and Improving Emotional Intelligence in LLMs via Psychologically Grounded Multi-Turn Dialogue

Large Language Models increasingly serve in emotionally sensitive roles, including mental health support, education, and crisis response, yet they lack a principled framework for assessing or improving Emotional Intelligence (EI). We introduce EiCAP, a unified, psychologically grounded six-layer EI taxonomy operationalized into two complementary resources. EiCAP-Bench is a multi-turn, one-vs-three forced-choice evaluation suite with 3,174 probes across 24 subcategories and cross-turn dependencies that reflect real conversational EI demands. EiCAP-SFT is a 152,820-dialogue supervision corpus aligned to the same taxonomy, enabling controlled, interpretable fine-tuning. Two key findings emerge. First, generic conversational supervised fine-tuning does not confer EI: fine-tuning on UltraChat yields no significant gain in any of the 24 subcategories, with a macro score of 24.6%, near the chance level of 25%. Second, applying EI-grounded LoRA, using approximately 0.8% of parameters, directly to Qwen-2.5-7B-Base achieves significant gains in all 24 subcategories, reaching a macro score of 75.33%, a gain of 51.7 percentage points over Base and 37.1 percentage points over Instruct. Crucially, an ablation shows that the UltraChat pre-stage is counterproductive, reducing performance by 21.4 percentage points: direct EI-grounded training is both necessary and sufficient.

21.
arXiv (math.PR) 2026-06-16

A non-asymptotic bound on the TV distance between a Wishart matrix and an appropriately scaled GOE matrix

arXiv:2606.16018v1 Announce Type: new Abstract: In this note, we prove a non-asymptotic version of a theorem by Bubeck, Ding, Eldan, and Rácz, showing that a Wishart matrix is close in total variation to an affine transformation of a GOE matrix. The proof mirrors the proof given by Bubeck et al., with some changes made to make it non-asymptotic.

22.
arXiv (CS.AI) 2026-06-19

A Systematic Evaluation of Black-Box Uncertainty Estimation Methods for Large Language Models

arXiv:2606.19868v1 Announce Type: new Abstract: Although large language models (LLMs) have shown strong capabilities across a wide range of tasks, their outputs often remain unreliable and may contain hallucinations, making uncertainty estimation (UE) essential for building trustworthy LLMs. In practice, many mainstream LLMs are only accessible through restricted APIs, where internal signals such as logits and hidden states are unavailable, making black-box UE especially important. However, existing work on black-box UE for LLMs remains fragmented in methodology and lacks a unified empirical comparison. To address this gap, we present a systematic review of black-box UE methods and organize them into five categories: verbalization-based, sampling-based, explanation-based, multi-agent, and hybrid methods. We further build a unified evaluation framework and benchmark 24 representative methods across 4 models and 4 dataset settings. Our results show that no single method consistently dominates across all settings. Nevertheless, methods that reason over and compare candidates in the answer space are generally effective, and hybrid methods that combine multiple uncertainty signals perform well under most conditions. By releasing the benchmark data and a unified evaluation framework, we aim to facilitate reproducible comparisons and support future research, while our empirical findings provide practical guidance for developing future black-box UE methods for LLMs.

23.
arXiv (CS.LG) 2026-06-16

Phase-Localized Curation Does Not Help: A Negative Result on Per-Phase Metric Selection for Demonstration Filtering

作者:

arXiv:2606.15064v1 Announce Type: new Abstract: Manipulation demonstrations have temporal phase structure, and a natural hypothesis is that demonstration-curation metrics should be applied within phases rather than globally. The idea is to segment each trajectory into phases, score each phase with the metric that is locally most informative, and then aggregate. This follows directly from prior work showing that a single global metric can be the best detector of a defect and yet the worst curator of the resulting policy. We test the per-phase hypothesis on three contact-rich LIBERO pick-and-place tasks with a controlled early-release structural defect, comparing phase-gated curation against the same metrics applied uniformly and against a strong single global metric. Across all three tasks and five random seeds per condition, phase-gated curation is never the best curation strategy, and it is the worst of the three on two of the three tasks (Task 1: 86.0 vs. 92.0 for global; Task 3: 22.7 vs. 48.0 for uniform). We trace the failure to a concrete mechanism. When the defect signal is concentrated in a single phase, rank-aggregating across phases dilutes that signal with uninformative scores from defect-free phases, selecting a worse demonstration subset than simply applying the defect-informative metric everywhere. We further show that the per-phase metric selection does not transfer across tasks, since no phase shares a winning metric between any two tasks, so the selection cannot be reused and must be re-derived per task from a noisy sweep. These results bound a plausible and previously untested method, and they argue that practitioners should prefer identifying a single defect-informative metric over decomposing curation by phase. We release the full pipeline, all metric implementations, and per-seed results.

24.
bioRxiv (Bioinfo) 2026-06-18

Predicting optimal growth temperatures of bacteria using learned structural information from a single protein

Temperature is a fundamental determinant of bacterial physiology and ecology. Optimal growth temperature (OGT) is highly variable across species, contributing to differences in where and when species are most likely to thrive. Although the OGTs for most bacteria remain unknown, the increasing availability of genomes from uncultivated and cultivated taxa has made it advantageous to build genomic, cultivation-independent models to infer OGT. However, pre-existing genomic models often lack the generalizability and mechanistic grounding required for robust inferences of OGT. We propose a novel framework for predicting bacterial OGT which uses learned protein structural signatures of thermal adaptation. We hypothesize that biophysical tradeoffs which dictate enzymatic functions across variable temperatures provide a more robust empirical basis for OGT prediction than broad genomic features. Our OGT-predicting model, ROSEATE, is based on a single gene, adenylate kinase (ADK), that encodes for a ubiquitous enzyme essential for energy homeostasis. ROSEATE uses high-dimensional latent space encoding via MSA Transformer, a protein language model which embeds ADKs in a manner which preserves biophysical information about embedded proteins. We show that the accuracy of the ROSEATE model is on par with other genome-based models, has a high degree of phylogenetic generalizability, and the ESM embeddings effectively capture key temperature-adaptive enzyme characteristics derived from AlphaFold structures. Because ROSEATE is based on analyses of a single ubiquitous protein, it can be used with metagenomic data to infer the community-level variation in bacterial OGTs. We demonstrate this feature of ROSEATE by reconstructing ADK sequences from over 500 environmental and host-associated metagenomes, successfully distinguishing community-wide thermal preferences across diverse habitats, from polar oceans to mammalian guts. By transitioning from genomic proxies to informationally dense protein structural features, this work provides an efficient, interpretable tool for predicting bacterial OGTs across taxa and whole communities.

25.
arXiv (CS.AI) 2026-06-19

The Autonomy Tax: Defense Training Breaks LLM Agents

arXiv:2603.19423v2 Announce Type: replace-cross Abstract: Large language model (LLM) agents increasingly rely on external tools (file operations, API calls, database transactions) to autonomously complete complex multi-step tasks. Practitioners deploy defense-trained models to protect against prompt injection attacks that manipulate agent behavior through malicious observations or retrieved content. We reveal a fundamental capability-alignment paradox: defense training designed to improve safety systematically destroys agent competence while failing to prevent sophisticated attacks. Evaluating defended models against undefended baselines across 97 agent tasks and 1,000 adversarial prompts, we uncover three systematic biases unique to multi-step agents. Agent incompetence bias manifests as immediate tool execution breakdown, with models refusing or generating invalid actions on benign tasks before observing any external content. Cascade amplification bias causes early failures to propagate through retry loops, pushing defended models to timeout on 99\% of tasks compared to 13\% for baselines. Trigger bias leads to paradoxical security degradation where defended models perform worse than undefended baselines while straightforward attacks bypass defenses at high rates. Root cause analysis reveals these biases stem from shortcut learning: models overfit to surface attack patterns rather than semantic threat understanding, evidenced by extreme variance in defense effectiveness across attack categories. Our findings demonstrate that current defense paradigms optimize for single-turn refusal benchmarks while rendering multi-step agents fundamentally unreliable, necessitating new approaches that preserve tool execution competence under adversarial conditions.