Leaderboards, eval harnesses and ablations — the contested business of deciding which model is actually better.
arXiv:2610.03448v1 Announce Type: cross Abstract: LLM agents increasingly screen tool outputs with small prompt-injection detectors, and teams choose among detectors by their scores on public benchma...
arXiv:2610.02388v1 Announce Type: new Abstract: Existing 3D large language models (LLMs) compromise on two fronts: they compress shapes into latent codebook indices or coordinate text, which removes...
arXiv:2608.17623v2 Announce Type: replace Abstract: Retinal diseases are a leading cause of irreversible vision impairment, making early and accurate diagnosis essential for effective treatment. Opti...
arXiv:2510.22276v4 Announce Type: replace-cross Abstract: Contrastive vision-language models have achieved remarkable progress through large-scale pretraining. Recent work has shown that removing Eng...
arXiv:2610.02270v1 Announce Type: cross Abstract: Medical vision-language model (VLM) evaluation is sensitive to workflow design, prompting strategy, and benchmark construction, yet most studies trea...
arXiv:2610.03599v1 Announce Type: new Abstract: High-fidelity 3D head avatars have reached near-photorealistic quality. While recent methods enable text-driven manipulation, they struggle to provide...
arXiv:2610.03252v1 Announce Type: new Abstract: Generative models of human-object interaction are bounded by the data that exists: everyday activities involve several objects, but most captured datas...
arXiv:2610.03248v1 Announce Type: new Abstract: Panoramic images provide a complete 360-degree field of view, enabling comprehensive scene understanding for embodied perception. However, heterogeneou...
The paper introduces VIGOR, a framework for zero‑shot visual generalization in model‑based reinforcement learning. VIGOR enforces latent‑space consistency through asymmetric weak‑to‑strong augmentations, dynamics‑level consistency, and encoder‑level stabilization, allowing the agent to handle unseen visual distractions while maintaining sample efficiency. Experiments on the DeepMind Control Suite and Robosuite demonstrate that VIGOR outperforms state‑of‑the‑art baselines, achieving significant gains in both environments.
MOF-VERIFY is a failure-aware agentic harness designed to improve hypothesis verification for metal‑organic frameworks (MOFs). It introduces a diagnostic benchmark with four task families—structural grounding, synthesis‑condition verification, evidence‑sufficiency verification, and MLIP‑based computational verification—to pinpoint failures in knowledge access, evidence acquisition, and reasoning. Guided by these diagnostics, MOF‑Verify addresses structural, literature, evidence‑sufficiency, and computational bottlenecks, achieving significant performance gains over direct inference and retrieval‑based baselines across multiple large language models.
ULTRADISCOVERY is a benchmark that tests abductive exploration in an epistemically open, interconnected universe. It features five domains where agents must revise an initial theory and predict outcomes of unseen cross‑domain interventions, with a $2 imes 2$ design manipulating representation openness and evidence alignment. Experiments show that agents struggle to compose evidence into transferable representations, especially when the representation is open, and that even with aids exact predictions are rare without large budgets.
Efficient Memory Crystallization (EMC) is a training‑free, test‑time framework that transforms each incoming graph domain into a compact, semantically faithful memory via a closed‑form distribution‑matching solution. By eliminating the need for auxiliary generative modules, EMC reduces redundant domain information and models inter‑domain dependencies through state‑evolving memories, yielding a tighter generalization error bound. Experiments show EMC outperforms state‑of‑the‑art baselines on non‑stationary graph data, cutting average runtime by 87.4% and GPU memory usage by 92.4% compared to a recent competitor.
arXiv:2610.02959v1 Announce Type: new Abstract: Recent text-to-image models have made substantial progress in photorealism, aesthetics, and text-image alignment. Yet visually appealing images can sti...
The paper introduces reverse Item Response Theory (IRT) for analyzing cancer drug‑response data, treating cancer types as latent subjects with resistance ability and drugs as items with evasion difficulty. Using 242,036 measurements from the GDSC2 database, the model estimates cancer‑type‑level in‑vitro resistance and drug‑level broad activity on a shared latent scale. Across four missingness regimes, reverse IRT outperforms simple averaging, achieving higher correlation (Δρ = +0.089 to +0.095 at 60% missingness) and the best Brier score among five methods, with stable classifications for 19 of 28 cancer types and 82% directional agreement in a cross‑platform PRISM replication.
arXiv:2610. 00009v1 Announce Type: new Abstract: Frequency-collapse attention [Zeris, 2026e] achieves large gains over standard dot-product attention by replacing the Q/K dot product with a bandpass-filtered inner product at a learned frequency.
The paper critiques current memorization audits for generative models, arguing that lacking a proper null distribution leads to misleading conclusions. It introduces two exact null tests—one permutation test for whole models and a calibrated test for single images—showing that many previously flagged memorizations disappear under these stricter controls. The authors also propose a scale‑restricted statistic based on the Intersection Euler Characteristic Profile to better detect distinct copied images.
The paper introduces a new class of selective state space models (SSMs) that move beyond traditional diagonal constraints by leveraging exact non‑Abelian group tracking and solvable affine transformation groups. It reports significant performance gains, including a 0.04‑degree dead‑reckoning error, high accuracy on Dyck‑2 and deep AST scope tracking tasks, and demonstrates that strict isometry is essential for lossless long‑range associative memory.
The paper investigates the reliability of attention‑head ablation as a causal inference tool in language models. Using GPT‑2 small, the authors find that a natural post‑projection zeroing method is almost uncorrelated with a corrected pre‑projection ablation and yields a completely different set of top‑5 important heads. They also show that binary accuracy can mask effects near performance floors or ceilings, whereas gold‑token log‑probability provides a graded signal. By employing a discovery/held‑out split and 1,000 matched random‑head and layer‑matched‑head controls, the corrected per‑head effect ranking remains highly stable (Spearman ρ = 0.974) and the top‑5 heads significantly outperform both control distributions (Monte Carlo p = 0.001). However, evidence for task specificity is weak on GPT‑2, and replication on DistilGPT‑2 confirms the intervention‑semantic and matched‑control findings.
SimplexUQ introduces the first benchmark and reproducible protocol for evaluating how conformal prediction wrappers allocate coverage across simplex‑valued predictions. The framework, called SimplexTasks‑12, combines six synthetic regimes and six real tasks (e.g., class probabilities, topic mixtures, spectral abundances) to compare existing wrappers on metrics such as marginal coverage, worst‑stratum coverage, max disparity, and computational cost. Empirical results show that no single wrapper consistently dominates, with Mondrian and BatchMVP performing best in different settings, and that removing predictor bias only partially mitigates disparity.
TrueMuse is a new benchmark designed to evaluate data attribution in text-to-music models. It consists of a controlled dataset created by fine‑tuning three diffusion‑based models on curated attribution samples, providing known attribution targets. The benchmark covers four settings—melodic structure, timbral characteristics, artist‑level style, and genre‑level patterns—across 133 attributes, 648 models, and 95,456 generated samples, and is used to assess existing black‑box attribution methods along several dimensions.