Search: “models”

Priced Guidance: Can Language Models Generate Future Research Ideas?

arXiv:2610.04976v1 Announce Type: new Abstract: We evaluate language models' capability to generate novel research ideas through the lens of compression. We aim to lower-bound the potentially tiny pro

6 Oct 2026

Safety

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Representation-Space MMD for Diffusion Language Models

DGX agent

arXiv:2610.06648v1 Announce Type: cross Abstract: We introduce a post-training method for diffusion language models (DLMs) that minimizes Maximum Mean Discrepancy (MMD) between generated and reference

safetyarxiv-cs-lg

6 Oct 2026

Local Ai

The Poisoned Conversation: Privacy-Leaking Watermarks in Unified Multimodal Models

arXiv:2610.05453v1 Announce Type: cross Abstract: Multimodal models are increasingly shifting toward unified architectures that understand and generate text, images, and other modalities within a shar

local-aiarxiv-cs-cv

6 Oct 2026

When Does Retrieval Help? A Study of In-Context Adaptation in Vision-Language-Action Models

arXiv:2610.05492v1 Announce Type: new Abstract: Vision-language-action (VLA) models have shown strong potential as generalist robot policies, but adapting them to unseen tasks often requires costly pa

6 Oct 2026

Controlling Polar Exposure to Delay Memorization in Diffusion Models

arXiv:2610.02780v1 Announce Type: new Abstract: Diffusion models can reach useful sample quality before copying training examples, but fast optimization can compress this generalization window by acce

Emergent Structure in the Marginal Attention Space of Language Models

arXiv:2610.03109v1 Announce Type: new Abstract: While representation similarity across independently trained language models is well-documented, how internal mechanics such as attention behave across

researcharxiv-cs-cl

NegT2IBench: When Negation Changes the Picture. A Polarity Benchmark for Text-to-Image Models

arXiv:2610.03084v1 Announce Type: cross Abstract: Text-to-image (T2I) models are judged by benchmarks that measure whether requested content appears, but these benchmarks largely overlook the compleme

Predicting and Repairing Merge Collapse in Large Language Models

arXiv:2610.03199v1 Announce Type: cross Abstract: Large language models fine-tuned from a shared base can be merged by averaging their task vectors, but some merges collapse far below the base model,

researcharxiv-cs-cl

Reliability Stress Tests and Decision-Time Routing for Chest X-ray Vision-Language Models

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 treat

Understanding Clinical Cognitive Dialogues Using Large Language Models

model-releasesarxiv-cs-cl

arXiv:2609.34125v2 Announce Type: replace Abstract: In-person cognitive assessment is both a test and an interaction. Clinicians explain tasks, repair misunderstandings, and adapt to patient responses

Safety

Debias Anything: Fairness with Diversity without Supervision in Diffusion Models

arXiv:2610.01815v1 Announce Type: new Abstract: Although diffusion models produce high-quality images, they also reproduce and amplify demographic imbalances in their training data. Debiasing their ge

safetyarxiv-cs-lg

DiVid: Diagnosing Dimension-Specific Diversity Collapse in Video Generation Models

arXiv:2610.01661v1 Announce Type: new Abstract: Despite remarkable progress, video generation models often produce highly similar outputs when repeatedly sampled from the same prompt, limiting their u

Higher-Order Molecular Grammars for Generative and Foundation Models in Chemistry

arXiv:2610.02186v1 Announce Type: cross Abstract: Molecular learning models are strongly shaped by their underlying representations. Yet standard sequential and graph formalisms struggle to explicitly

Memorizon: Training World Models Beyond Their Context Window

arXiv:2610.00544v1 Announce Type: new Abstract: Streaming world models should render a place consistently across repeated visits. Directly supervising such revisits requires training samples that capt

Tutorials

RelICL: Training-free Relational Learning with Tabular Foundation Models

arXiv:2610.01725v1 Announce Type: new Abstract: Tabular foundation models achieve state-of-the-art performance on single-table tasks without any training. Recent work suggests that they are also well-

tutorialsarxiv-cs-lg

ROWBench: Do Video Models Render What the Program Specifies?

arXiv:2610.02205v1 Announce Type: new Abstract: Programmable world models separate executable dynamics from visual generation, offering a promising foundation for next-generation game engines. However

The Null Is the Hard Part: Exact Tests for Memorization in Generative Models

arXiv:2610.00251v1 Announce Type: new Abstract: Memorization audits of generative models read similarity scores against thresholds, with no null distribution, and the conclusions they support can be w

UnifiedAttack: Evaluating the Safety of Large Multimodal Models in Synergistic Harmful Image-Text Generation

arXiv:2610.00341v1 Announce Type: cross Abstract: As Large Multimodal Models (LMMs) transition toward natively unified architectures, evaluating their safety in synergistic harmful image-text generati

Role-guided Speaker Deletion Verification in Clinical Psychiatry Speech Recordings with Audio Language Models

arXiv:2609.38491v1 Announce Type: new Abstract: Clinical research in psychiatry increasingly relies on large scale collection of spoken language data to identify acoustic and linguistic biomarkers. Ye

Self-Repulsive Sampling for Diffusion Language Models

arXiv:2609.39560v1 Announce Type: new Abstract: Sampling several responses and voting over their answers can improve a language model's accuracy, but repeated answers limit the benefit of additional s

Hardware

Sparse-WAM: Accelerating World Action Models via Action-Guided Sparse Imagination

arXiv:2609.38984v1 Announce Type: cross Abstract: World-action models (WAMs) leverage pretrained video models to improve generalization in robot control by jointly predicting future visual states and

hardwarearxiv-cs-ai

Agents

Thinking Outside the Box: Can Language Models Rely on External Guidance Selectively?

arXiv:2609.39578v1 Announce Type: new Abstract: Agent harnesses often improve language models with human-designed workflows, but as models grow more capable, unreliable guidance can increasingly const

agentsarxiv-cs-cl

BarcodeMAE+: Rethinking Masked Pretraining and Global Representations for DNA Barcode Foundation Models

arXiv:2609.35877v1 Announce Type: cross Abstract: Many DNA foundation models are pretrained by masking parts of a sequence and asking the model to reconstruct them. Standard masked pretraining exposes

researcharxiv-cs-lg

Beyond Symmetric Agents: Cognitive Diversity and Multi-Agent Debate in Small Language Models

arXiv:2609.35875v1 Announce Type: new Abstract: Multi-agent debate (MAD) reportedly improves reasoning and factuality over single-model inference, but prior work treats agents as symmetric peers, leav

Cross-Entropy Guided Routing in Mixture-of-Experts Large Language Models

arXiv:2609.37751v1 Announce Type: new Abstract: Sparse mixture-of-experts (MoE) large language models scale model capacity by routing each token to a small subset of experts. Their routers are regular

FastGuide: Accelerating Reward Guidance for Diffusion Large Language Models

arXiv:2609.36202v1 Announce Type: new Abstract: Gradient-based reward guidance provides a flexible way to use downstream reward models to control masked diffusion language models at inference time. Ho

researcharxiv-cs-cl

Applications

Support-Set Target Leakage in Relational Foundation Models during In-Context Learning: Model Dependence and Evaluation Reliability

arXiv:2609.36417v1 Announce Type: new Abstract: Relational in-context learning (ICL) uses labeled support examples and their linked relational context to predict labels for new queries. This creates a

applicationsarxiv-cs-ai

When Models Don't Manipulate Manifolds: The Geometry of a Comparison Task

arXiv:2609.37680v1 Announce Type: cross Abstract: One of the current premises of mechanistic interpretability research is that detailed accounts of the geometry of neural network representations can t

Commutator Memory: Sparse, Path-Local Reading and Steering in Language Models

arXiv:2609.34348v1 Announce Type: new Abstract: Gradient updates on different data generally do not commute: training a language model on two data sources in opposite orders gives different weights, e

Identical Runs, Different Results: Benchmarking AI Coding Agents on Open-Weight Models

arXiv:2609.33812v1 Announce Type: cross Abstract: Repeated runs of the same coding agent are known to give different benchmark scores. We ask what that variation means for a team running an agent on i

Multimodal LLMs Outperform Pathology Foundation Models in Cross-Domain Histological Similarity

arXiv:2609.32876v1 Announce Type: new Abstract: State-of-the-art pathology foundation models, trained on millions of histology tiles, can fail to preserve tissue similarity when comparisons cross slid

REMEDY: How Far Is Video Generation from Medical Education World Models?

arXiv:2609.32460v1 Announce Type: new Abstract: Recent video generation models produce realistic videos and show potential as a foundation for world models. These advances create opportunities for gen

Separating personal from population gains when calibrating EEG foundation models for new users

arXiv:2609.34801v1 Announce Type: new Abstract: Foundation models are increasingly adapted to individual users, but an apparent personalization gain can simply reflect a stronger population model. Thi

researcharxiv-cs-lg

Typed Decision Models: An Early Evidence Audit and Evaluation Checklist

arXiv:2609.32160v1 Announce Type: cross Abstract: Typed decision models (TDMs) return probability distributions over caller-defined options without generating text. TypeSafe released Jev, a commercial

VisionPsy-Nano: Improving Accuracy, Efficiency, and Reliability in On-Device Vision-Language Models

arXiv:2609.31746v1 Announce Type: new Abstract: Sub-billion-parameter Vision-Language Models are increasingly viable for on-device deployment, yet compact model size alone does not guarantee usability

Evaluating Sycophancy in Chinese Large Language Models on Factual Questions Derived from Online Search Queries

model-releasesarxiv-cs-cl

arXiv:2609.30986v1 Announce Type: new Abstract: As large language models increasingly mediate information access, factually accurate and independent answers are critical. However, these models can exh

Neural State Prediction: Obstructing Shortcut Learning in EEG Foundation Models

arXiv:2609.31167v1 Announce Type: new Abstract: EEG foundation models increasingly use masked prediction to learn from unlabeled recordings, but optimizing this objective does not ensure transferable

Training Graph Foundation Models on The Web Graph

arXiv:2609.30894v1 Announce Type: new Abstract: We introduce Acacia, a graph foundation model, trained on the web graph. Acacia (i) supports arbitrary feature dimensionalities and semantics without ad

researcharxiv-cs-ai

UltraG-Bench: A Multi-task Benchmark for assessing Large Vision-Language Models on Pixel-level Evidence Grounding in Ultrasound

arXiv:2609.30928v1 Announce Type: cross Abstract: Ultrasound is one of the most widely used medical imaging modalities, and recent large vision-language models(VLMs) have shown increasing capabilities

VLALight: Lightweight Vision-Language-Action Models for Emergency-Aware Traffic Signal Control

arXiv:2609.30709v1 Announce Type: cross Abstract: Traffic signal control (TSC) is essential for mitigating urban congestion. Recent advances in vision-language models (VLMs) enable richer interpretati

PROOF: Profiling Reliability of Object-Level Facts in Large Language Models

arXiv:2609.29504v1 Announce Type: cross Abstract: Aggregate factuality scores hide where a language model succeeds, which relations it confuses, and whether an answer survives innocuous changes to the

25 Sept 2026

Benchmarking Hyperspectral Foundation Models for Hyperspectral Unmixing

arXiv:2609.28283v1 Announce Type: new Abstract: Several foundation models dedicated to hyperspectral images have recently been made available. These models are trained on large unlabeled datasets and

Beyond Poetry: Can Large Language Models Generate Classical Arabic Maqamat?

arXiv:2609.28245v1 Announce Type: cross Abstract: Large language models (LLMs) have shown strong performance in creative text generation, yet their ability to produce culturally grounded and stylistic

Evaluation of pre-trained models for pedagogical assessment of novel AI-assisted educational questions

arXiv:2609.27749v1 Announce Type: new Abstract: The surge in AI-assisted generation of educational materials has outpaced our capacity to validate their pedagogical quality. Automated evaluation using

researcharxiv-cs-ai

Parameter Importance-Driven Continual Learning for Foundation Models

arXiv:2511.15375v2 Announce Type: replace-cross Abstract: Domain-specific post-training often causes catastrophic forgetting, making foundation models lose their general reasoning ability and limiting

PRISM-VLM: A Multi-Axis Discriminative Benchmark for Compact Vision-Language Models

model-releasesarxiv-cs-cl

arXiv:2609.27395v1 Announce Type: new Abstract: Compact vision-language models (VLMs) now power a growing share of multimodal applications. The benchmarks used to compare them, however, inherit a fron

Safety

Double Descent and Malign Overfitting in Diffusion Models

arXiv:2609.26392v1 Announce Type: new Abstract: Conventional wisdom in deep learning holds that overparameterization---having more parameters p than training samples n---is benign: larger models gener

safetyarxiv-cs-lg

23 Sept 2026

Pinocchio: Fast Uncertainty Estimates for Black-Box Language Models