Hugging Face Daily Papers
500 articles archived · Visit source ↗ · RSS
-
Hugging Face Daily Papers research 12d ago
Plausible but Not Valid: A Psychometric Audit of LLMs as Synthetic Survey Respondents
Abstract Large language models replicate broad psychometric trends in synthetic survey responses but fail to match human joint distributions, reliability, and mediation structures, making them unsuitable replacements for real respondents. Generated by…
6 -
Advancing Open and Reproducible Relational Learning: RelArena-α, TabPFN-Rel and RPI
Abstract Prior Labs released open-source tools including a unified relational benchmark framework, a TabPFN-based relational model, and a model-agnostic predictive interface to advance reproducible relational learning. Generated by thinkingmachines/Inkling-Small This first…
21 -
GRNEdit: Efficient General Video Editing from a New Binary-Evidence Perspective in Generative Refinement Networks
Abstract GRNEdit is a lightweight two-stage framework that models video editing intent via binary semantic decisions and source evidence, achieving strong results with minimal parameters. Generated by thinkingmachines/Inkling-Small Instruction-based general video editing seeks…
16 -
MOSS-VL Technical Report
Abstract MOSS-VL is an open vision-language model family enabling real-time interaction by attending to vision via gated cross-attention during generation, using a synthesized interaction corpus and staged curriculum to achieve strong streaming performance with reduced…
25 -
A Plug-and-Play 2D Motion Interface for Real-World Motion Language Models
Abstract A plug-and-play 2D motion interface allows pretrained motion language models to process 2D inputs without retraining, improving real-world applicability. Generated by thinkingmachines/Inkling-Small Motion Language Models (MoLMs) typically understand human motions by…
22 -
How Do Agents Fail on AutoResearch: End-to-End Diagnostic Evaluation on 100 Real-World Frontier Research Tasks
Abstract Autonomous research agents evaluated across the full scientific lifecycle reveal a pervasive lack of metacognitive self-correction, motivating a new benchmark and failure taxonomy. Generated by thinkingmachines/Inkling-Small AI has long assisted scientific research, but…
8 -
Large Discovery Models: Empirically-grounded Model-Based Open-Ended Search
Abstract A recurrent Large Discovery Model couples generative proposal with a Bayesian non-parametric reward surrogate to guide uncertainty-aware search across molecules, proteins, and programs. Generated by thinkingmachines/Inkling-Small Scientific discovery often involves…
14 -
TRACE-Bench: Decomposing and Diagnosing Multi-Reference Image Generation
Abstract This work proposes a compositional operator framework and TRACE-Bench to diagnose multi-reference image generation capabilities across atomic operations. Generated by thinkingmachines/Inkling-Small Despite recent advances in unified multimodal models for multi-reference…
22 -
WorldRover: A Scalable Synthetic Video Data Engine for World Exploration with Rich Annotations
Abstract WorldRover is a synthetic data engine that generates long-range, richly annotated video sequences with depth, camera motion, and tracking signals to support training models for coherent world exploration. Generated by thinkingmachines/Inkling-Small Learning to generate…
32 -
Prototype-Rectified Iterative Self-supervised Manifold Denoising under Severe Acoustic Shift
Abstract PRISM is a fast, training-free test-time adaptation method that reverses low-rank affine noise distortions in audio-text models using frozen text prototypes and geometric corrections. Generated by thinkingmachines/Inkling-Small Audio-Text Foundation Models (ATMs) fail…
33 -
Learn What's Left, Not What's Mastered: Saturation Aware Advantage Reweighting for Multi-Reward Policy Optimization
Abstract SA-MRPO independently standardizes multi-objective rewards and adaptively discounts saturated objectives to redirect optimization toward under-optimized goals. Generated by thinkingmachines/Inkling-Small Reinforcement learning (RL) with group-relative advantages has…
23 -
DumpsterCluster: From Dumpster Diving to Serving LLaMA-70B on $60 GPUs
Abstract Retired GPUs can form low-cost clusters for LLM inference, but their economic and environmental viability depends heavily on local electricity prices and carbon intensity. Generated by thinkingmachines/Inkling-Small As AI datacenters retire functional GPUs, vast…
9 -
HiFi-BRep: High-Fidelity Latent Representation for Robust B-Rep Generation
Abstract HiFi-BRep improves B-Rep synthesis by using a topology-aware encoder and a single-stage decoder that jointly predicts geometry and topology with differentiable validity constraints. Generated by thinkingmachines/Inkling-Small Boundary representation (B-Rep) generation…
31 -
Drive, Pack, Fly: The Travelling Thief Problem with Drone
Abstract The Travelling Thief Problem with Drone jointly optimizes ground routing, drone synchronization, and item selection to maximize profit, using mixed-integer programming, metaheuristics, and attention-based deep reinforcement learning with a hybrid refinement approach.…
33 -
AnyTalk: Speech Animation for Arbitrary Characters Leveraging a Video Generation Model
Abstract AnyTalk generates 3D speech animations for arbitrary characters without animation data by adapting video diffusion models via character-specific fine-tuning and optimizing blendshape parameters from synthesized talking-head videos, with a distilled real-time variant.…
9 -
Prior Audit-Repair Context Shifts LLM Verifier Thresholds Toward Leniency
Abstract Prior audit and repair episodes in context reduce false alarms by shifting decision thresholds rather than discrimination, with repair content and audit verdict complementarily affecting different model families. Generated by thinkingmachines/Inkling-Small Automated…
35 -
Gathered, Not Admitted: How Attention Brings a Latent Variable into Verbalizable Form
Abstract In language models, flexible reuse demands attention-mediated gathering at a mid-depth window to make latent variables readable, without a selective gate, and readout measures poorly reflect actual use. Generated by thinkingmachines/Inkling-Small Language models hold…
29 -
HarnessEval-W: Agentifying the Evaluation of Visual Worlds
Abstract HarnessEval-W uses hierarchical sub-agents to decompose world-model evaluations into verifiable reasoning chains that justify scores with transparent evidence. Generated by thinkingmachines/Inkling-Small A benchmark should deliver more than a scalar score: what makes an…
29 -
Improving the matrix multiplication exponent with modern optimization and AlphaEvolve
Abstract Refinements to combination loss analysis via reformulated optimization, machine learning-based algorithms, and AlphaEvolve yield an improved upper bound on the matrix multiplication exponent. Generated by thinkingmachines/Inkling-Small The current best bounds on the…
35 -
An Empirical Study of Training Pixel-Space Text-to-Image Diffusion Models
Abstract Researchers propose a latent-to-pixel training strategy that accelerates convergence and improves inference speed for large-scale pixel-space diffusion models. Generated by thinkingmachines/Inkling-Small This paper investigates an increasingly important topic in…
16 -
PACE-Bench: Benchmarking Physics Adaptation via Code Evolution in Dynamic Environments
Abstract PACE-Bench evaluates self-evolving agents on physics adaptation tasks requiring iterative code redesign after environmental mutations, revealing that simulator-grounded reflection outperforms unverified self-revision but mechanism redesign remains a major bottleneck.…
30 -
VideoGAIA: A Benchmark for General AI Assistants on Agentic Video Understanding
Abstract VideoGAIA introduces a multi-turn, tool-augmented benchmark that evaluates agentic video understanding for advanced multimodal models through complex real-world tasks. Generated by thinkingmachines/Inkling-Small Video understanding is a fundamental task for evaluating…
4 -
UI-Mate: Advancing Open-Weight Foundation GUI Agents with In-Context Demonstrations
Abstract UI-Mate is a foundation GUI agent that uses environment-grounded training and in-context demonstration learning to improve reliability on long-horizon office tasks, achieving state-of-the-art results on computer-use benchmarks. Generated by…
30 -
Position: AI Agents in Scientific Teams Should Be Studied as Human-Agent Systems
Abstract Scientific collaboration with AI agents requires studying human-agent pairs to avoid risks like reduced inquiry diversity and to foster synergistic discovery. Generated by thinkingmachines/Inkling-Small Large language model-based agents are increasingly deployed as…
7 -
When Context Bites: Detecting RAG Poisoning via Document-Level Attention Collapse
Abstract D-SCAN detects retrieval poisoning by monitoring attention collapse dynamics in language model generations. Generated by thinkingmachines/Inkling-Small Retrieval-augmented generation (RAG) is indispensable for enhancing large language models. However, RAGs are…
23 -
ConceptFormer: Learning Adaptive Latent Concepts for Query-Document Alignment in Visual Document Retrieval
Abstract ConceptFormer learns continuous latent concept representations to bridge visual evidence and semantic relevance for visual document retrieval without relying on text intermediates or raw visual annotations. Generated by thinkingmachines/Inkling-Small Visual document…
13 -
Ventor-QTest: Threat-Model-Driven Verification of Vendor-Hosted LLM APIs
Abstract Ventor-QTest audits hosted open-weight model APIs via repeated and long-sequence black-box probes, measuring average and extreme fidelity loss to detect degradation in long-horizon agentic performance. Generated by thinkingmachines/Inkling-Small As large language models…
35 -
ENTLORE: A Graph-Grounded Benchmark for Latent Organizational Reasoning in Enterprise Question Answering
Abstract ENTLORE is a benchmark framework that evaluates enterprise question answering by requiring recovery of implicit organizational relations across routine documents, revealing that even with gold sources many latent reasoning questions remain unanswered. Generated by…
7 -
ClawGym II: Exploring Black-Box RL on Agent Harness
Abstract A unified black-box reinforcement learning framework enables stable, scalable optimization of general agents through complex harnesses via sandbox execution, trajectory reconstruction, and mix-harness training. Generated by thinkingmachines/Inkling-Small Agent harnesses…
32 -
R^3-Bench: LLMs Struggle with Resource-Rational Reasoning under Shared Budgets
Abstract R³-Bench reveals that shared computation budgets cause reasoning agents to underperform relative to their single-problem capabilities across math, coding, and abstract reasoning tasks. Generated by thinkingmachines/Inkling-Small In cognitive science, resource…
7 -
GenRouter: Unified Workflow Routing for Agentic Image Generation
Abstract GenRouter is a unified routing framework that adaptively directs prompts to optimal agentic image-generation workflows, cutting costs and latency while improving visual alignment and enabling continuous self-evolution. Generated by thinkingmachines/Inkling-Small The…
34 -
VibeWorlding: Can Multimodal Agents Construct 3D Open Worlds End-to-End?
Abstract A unified framework benchmarks and trains multimodal agents that infer intent, plan 3D scenes, invoke tools, and reflect on feedback, revealing that reinforcement learning improves open-source models beyond closed-source frontiers. Generated by…
14 -
Understanding Cognition-Induced Risks in Agentic AI Systems
Abstract Agentic systems built on large language models pose escalating risks to human agency and autonomy across physical, social, and self-referential cognitive levels, requiring targeted mitigation strategies. Generated by thinkingmachines/Inkling-Small Frontier agentic…
38 -
MegaParts: Scaling Part-Aware 3D Object Generation to 300 Parts via Token-Efficient Autoregressive Modeling
Abstract MegaParts scales part-aware 3D generation via token-efficient vector-quantized part tokens and structured autoregressive sequence modeling with long-context training. Generated by thinkingmachines/Inkling-Small Part-aware 3D object generation is essential for graphics…
21 -
Agentic Transaction: Towards ACID-Compliant Agent Systems
Abstract An ACID-compliant framework for agentic transactions introduces semantic guarantees to ensure reliable, isolated, and durable execution of long-horizon LLM agent workflows. Generated by thinkingmachines/Inkling-Small Large language model (LLM) agents are evolving from…
4 -
NaviDC-OCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
Abstract NaviDC-OCR is a unified vision-language framework that integrates deformation-aware learning, adaptive layout sampling, and decoupled content-structure training to improve document parsing accuracy and structural reasoning. Generated by thinkingmachines/Inkling-Small…
17 -
Who Speaks Matters: Authority-Aware Multi-View RAG over Italian Parliamentary Proceedings
Abstract ParliamentRAG is a retrieval-augmented generation system for Italian parliamentary records that uses topic-dependent speaker authority to retrieve expert perspectives and generate faithful, multi-perspective summaries. Generated by thinkingmachines/Inkling-Small…
24 -
DFM Mimir v1: An Open HRM Delivering Frontier Performance at 1B Parameters Using Only Permissible Post-Training Data
Abstract Mimir v1 is a 1-billion-parameter Hierarchical Reasoning Model trained solely on permissible data that achieves competitive English results and state-of-the-art Danish performance across multiple benchmarks. Generated by thinkingmachines/Inkling-Small Current large…
20 -
Agents Catching Agents: Shortcut Cascades and Benchmark Gaming in Clinical Multi-Agent Systems
Abstract Multi-agent clinical committees are vulnerable to socially plausible shortcuts rather than isolated cues, and only independent referee oversight reliably detects adoption. Generated by thinkingmachines/Inkling-Small Clinical decision support is moving toward committees…
35 -
A Pathway to General-Purpose Scientific AI: Multimodal Comprehension of Scientific Images
Abstract The ALD/E-ImageMiner benchmark and ICDAR 2026 competition advance machine interpretation of scientific figures through tasks spanning visual reading, domain reasoning, and evidential justification, proposing long-term goals for verifiable multimodal scientific AI.…
30 -
Intern-S2-Mobius: Foundation Model with Decoupled Knowledge and Reasoning
Abstract Mobius-v0 separates global memory storage from iterative reasoning modules to improve knowledge compression and inference efficiency, yielding comparable performance with less training data and faster inference. Generated by thinkingmachines/Inkling-Small We introduce…
24 -
Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction
Abstract GAS improves multimodal understanding by using generation as auxiliary supervision via next embedding prediction and a decoupled mixture-of-transformers architecture, with no inference overhead. Generated by thinkingmachines/Inkling-Small While Multimodal Large Language…
19 -
SPARGen: Unifying Spatial Perception and Reasoning through Native Multimodal Generation
Abstract SPARGen unifies 3D reconstruction, dense correspondence, and spatial reasoning into a single instruction-conditioned multimodal generative model that jointly learns shared spatial representations. Generated by thinkingmachines/Inkling-Small Spatial perception and…
31 -
Self-Supervised Visual On-Policy Distillation
Abstract Self-supervised visual on-policy distillation improves small vision-language models by distilling from original images into strongly augmented student views without privileged annotations or larger teachers. Generated by thinkingmachines/Inkling-Small Visual on-policy…
5 -
SimpleOPD: Simple Tokenizer-Agnostic On-Policy Distillation for Long-Context Reasoning
Abstract On-policy distillation from a long-context reasoning teacher to short-context students improves mathematical proof reasoning and generalizes to science benchmarks by aligning token spans, constraining length growth, and stabilizing training. Generated by…
26 -
LittleLearner: Language Models Under Pedagogically Controlled Knowledge Exposure
Abstract A curated elementary-grade pretraining corpus and 5B-parameter model create a controlled sandbox for studying knowledge acquisition, representation, and bounded capability growth via post-training and in-context learning. Generated by thinkingmachines/Inkling-Small…
14 -
UniProbe: A Learnable Token-Level Hallucination Detector for Large VLMs using Multi-Structural Internal Representations
Abstract UniProbe is a lightweight learnable detector that uses a directed graph and alternating GNN, ViT, and GRU modules to identify hallucinated tokens in frozen large vision-language models, enabling real-time resampling during generation. Generated by…
22 -
Forecast Collapse in Time-Series Foundation Models
Abstract Forecast collapse in hourly equity return prediction stems from low predictability and per-series objectives, and the proposed CalibRank objective balances calibration and ranking to restore cross-sectional structure. Generated by thinkingmachines/Inkling-Small When…
28 -
Second Thought: Reasoning in Parallel as LLM Agents Act and Observe
Abstract Second Thought is a training-free framework that runs auxiliary reasoning branches in parallel during agent action-observation waits to reduce sequential decoding and turn counts without harming accuracy. Generated by thinkingmachines/Inkling-Small LLM agents in the…
35 -
Latent On-Policy Self-Distillation
Abstract Latent On-Policy Self-Distillation learns privileged teaching context end-to-end from experience to provide dense token-level supervision, improving agent performance and efficiency. Generated by thinkingmachines/Inkling-Small Enabling agents to learn from experience…
10