On-policy distillation (OPD) is increasingly used to improve large language model reasoning, but its training dynamics remain poorly understood. We characterize the trajectory of…
HuggingFace Papers
Video world models that maintain 3D spatial consistency across generated frames typically rely on explicit point cloud memory constructed in RGB space. This design is both…
HuggingFace Papers
Agent systems increasingly use textual skills to encode reusable task procedures, but injecting these skills into the prompt at every step incurs substantial context overhead and…
HuggingFace Papers
While recent text-guided video editing models excel at elementary tasks (e.g., style transfer, object insertion), real-world user requests are highly compositional. A single…
HuggingFace Papers
Conventional LLMs keep the full KV cache loaded during decoding, causing a severe GPU memory bottleneck for ultra-long context serving. In this report, we propose Lookahead Sparse…
HuggingFace Papers
We examine whether human psychometric questionnaires can serve as reliable tools for characterizing and predicting LLM behavior in everyday user interactions. We analyze eight…
HuggingFace Papers
We present Echo-Memory, a controlled study of memory mechanisms in action-conditioned world models. These models generate multi-segment videos from a first frame, text prompt, and…
HuggingFace Papers
Vision-language model (VLM) agents are increasingly deployed in interactive game environments. Yet game benchmarks for VLM agents typically report a single first-attempt score per…
HuggingFace Papers
Long-context language model inference is bottlenecked by memory, as the KV cache grows with context length. Recent techniques to compress the KV cache fall short: they either…
HuggingFace Papers
Real-time video restoration (VR) for live streams requires high-resolution outputs under strict per-frame latency constraints. Existing one-step diffusion-based VR models remain…
HuggingFace Papers
World-action models have emerged as a promising paradigm for robot manipulation, jointly modeling visual scene dynamics and actions to inject physical priors into policy learning.…
HuggingFace Papers
LLM agents increasingly rely on external inference conditions: prompts, tools, memory, SOPs, skills, and harness feedback. These assets can improve task execution without changing…
HuggingFace Papers
Reward models (RMs) provide critical feedback signals for LLM post-training, notably in reinforced fine-tuning (RFT) and reinforcement learning (RL) pipelines. However, current…
HuggingFace Papers
While Omni-modal Large Language Models (OLLMs) have demonstrated impressive capabilities in jointly processing audio and visual streams, their ability to strictly adhere to…
HuggingFace Papers
We present DEI: Diversity in Evolutionary Inference, a distributed Quality-Diversity (QD) search framework that assigns heterogeneous large language models (LLMs) as mutation…
HuggingFace Papers
Whisper, a widely adopted ASR model, is known to suffer from hallucinations - coherent transcriptions generated for non-speech audio entirely disconnected from the input. We…
HuggingFace Papers
Deep research agents have demonstrated remarkable capabilities in complex information-seeking tasks, yet this power comes at a steep computational cost. Driven by accuracy-focused…
HuggingFace Papers
Speech-based large language models are typically constrained to spoken replies, which limits their user-facing outputs to what can be verbalized and suppresses text-native…
HuggingFace Papers
Retrieval-augmented QA pipelines often route retrieved passages through an LLM rewriter before a smaller reader, lifting F1 by tens of points on multi-hop benchmarks; this gain is…
HuggingFace Papers
Reinforcement learning with verifiable rewards (RLVR) has become a leading paradigm for improving the reasoning ability of large language models through outcome-based supervision.…
HuggingFace Papers
Muon improves training efficiency over Adam in large language-model training by about two times, but the local geometric source of this advantage remains unclear. Our work takes a…
HuggingFace Papers
World Action Models (WAMs) extend robot policy learning by incorporating future prediction as an additional training objective, encouraging the policy to encode task-relevant…
HuggingFace Papers
Linear activation steering has gained popularity as a simple and empirically effective way to control language model behavior. More recently, spherical steering paradigms have…
HuggingFace Papers
Text-to-image models rely on text prompts as their primary interface to human intent. Prompts are encoded by a text encoder into embeddings that condition the image generation…
HuggingFace Papers
Deep Research (DR) has emerged as a new agentic paradigm to tackle complex, open-ended research tasks, demanding systems that can iteratively frame problems, acquire evidence,…
HuggingFace Papers
On-policy distillation (OPD) has become a central post-training tool for large language models (LLMs), providing dense per-token teacher supervision along the student's own…
HuggingFace Papers
Vision-Language-Action (VLA) models are emerging as a promising paradigm for robotic manipulation, enabling general-purpose policies trained from large corpora of demonstrations…
HuggingFace Papers
Vision Transformers operate on fixed patch grids, which can introduce phase-dependent instability for dense prediction: changing the patch partition can change the token evidence…
HuggingFace Papers
Recent video-based world models have made pixel-space environments interactive at the camera level: users can navigate viewpoints while the model generates coherent visual…
HuggingFace Papers
Chain-of-Thought (CoT) improves the performance of Large Language Models (LLMs) and has been extended to Multimodal Large Language Models (MLLMs). More recent work further moves…
HuggingFace Papers
AI evaluation results are produced at scale but reported inconsistently across leaderboards, model cards, benchmark papers, and company blogs. The cost is interpretive: readers…
HuggingFace Papers
We present SigmaScale, a method for learning auxiliary scaling matrices S to aid truncated Singular Value Decomposition (SVD) based Large Language Model (LLM) compression. Instead…
HuggingFace Papers
Understanding what generative models retain from training data remains challenging, with implications for copyright and privacy. Beyond verbatim reproduction, models can encode…
HuggingFace Papers
Existing scientific relation extraction benchmarks mainly target domains such as computer science, where entities are tasks, methods, datasets, materials, or metrics. This leaves…
HuggingFace Papers
This paper explores agentic 3D spatial understanding, i.e., MLLM agents performing 3D reasoning through tool use. Existing methods often misuse tools and exhibit biased tool…
HuggingFace Papers
Agent benchmarks score submissions with outcome verifiers that are typically hand-written and brittle, leaving them open to reward hacking. We audit 1,968 tasks across five…
HuggingFace Papers
Standard transformers apply self-attention uniformly at every layer and token, regardless of whether the input requires dynamic cross-token interaction. We propose CHIAR-Former…
HuggingFace Papers
Long-horizon agentic tasks pose a fundamental credit assignment challenge for outcome-base reinforcement learning: trajectory-level rewards verify final correctness but provide…
HuggingFace Papers
Medical agent systems are increasingly expected to support interactive clinical decision making rather than only static question answering. In such settings, effective agents must…
HuggingFace Papers
Latent visual reasoning (LVR) inserts supervised latent tokens between perception and answer generation in vision-language models (VLMs). The field uses alignment between these…
HuggingFace Papers
Large language models are increasingly evaluated by other models, raising a natural question: can a model predict how a judge will score its own output? We find that the ability…
HuggingFace Papers
Equipping Large Language Models (LLMs) to execute reliable multi-step workflows has become a central challenge in artificial intelligence. Despite recent advances in LLMs' agentic…
HuggingFace Papers
Recent progress in robot manipulation has been largely driven by learning from large-scale demonstrations. For humanoid robot loco-manipulation tasks, however, existing data…
HuggingFace Papers
Mixture-of-Experts (MoE) is now the dominant architecture for frontier language models, yet it requires all expert parameters to be loaded in memory, making it less preferable for…
HuggingFace Papers
Large language models (LLMs) offer a promising approach to machine translation (MT) for extremely low-resource languages by incorporating linguistic resources through in-context…
HuggingFace Papers
Cross-view geo-localization estimates the geographic location of a ground image by matching it against an aerial image database. Existing methods tackle this through either…
HuggingFace Papers
We introduce EMMA, a physics-informed multimodal framework that recovers all identifiable dynamical parameters of a system directly from raw video, audio, and image-based…
HuggingFace Papers
Enterprise property graphs vary widely in schema structure, internal terminology, domain assumptions, governance constraints, and user interaction patterns. A deployment-relevant…
HuggingFace Papers
Reflexion-style agents rely on self-generated reflections as memory, implicitly assuming that agents can accurately diagnose their own failures. We show that this assumption can…
HuggingFace Papers
Passive long-wave infrared (LWIR) hyperspectral imaging under a standoff geometry depends on atmospheric absorption and emission, as well as reflected radiance, thus making…
HuggingFace Papers