Sumi: Open Uniform Diffusion Language Model from Scratch
Diffusion models have become a promising alternative to autoregressive models. Among these, uniform diffusion language models (UDLMs) permit any token to be updated at any step, in...
Research Papers
Latest AI/ML research from ArXiv and Papers with Code
Learning User Simulators with Turing Rewards
Learning to simulate human users in interactive settings could advance the training of agent assistants, evaluation of personalization systems, research in the social sciences, and...
TurboServe: Serving Streaming Video Generation Efficiently and Economically
Streaming video generation is emerging as a new serving workload in which users interact with long-lived sessions that generate video progressively, chunk by chunk. Unlike offline ...
OpenRath: Session-Centered Runtime State for Agent Systems
Modern agent systems often suffer from fragmented runtime state: transcripts, tool effects, memory events, workspace placement, branch provenance, and replay evidence are recorded ...
Learning from Your Own Mistakes: Constructing Learnable Micro-Reflective Trajectories for Self-Distillation
Self-distillation improves reasoning in large language models by using the model's own rollouts as training signal, typically through implicit logit-level alignment that minimizes ...
Qwen-RobotManip Technical Report: Alignment Unlocks Scale for Robotic Manipulation Foundation Models
Foundation models in language and multimodality achieve strong generalization by aligning heterogeneous data under a unified formulation and training at scale. In this report, we i...
Comparing Linear Probes with Mahalanobis Cosine Similarity
Linear probes are widely used in interpretability research and often compared by cosine similarity. The Mahalanobis cosine similarity (MCS) between two directions, which reweights ...
Playful Agentic Robot Learning
Current agentic robot systems can write executable Code-as-Policy programs, observe feedback, and revise behavior across multiple attempts, but they remain largely task-driven: reu...
Bag of Dims: Training-Free Mechanistic Interpretability via Dimension-Level Sign Patterns
We show the standard basis of transformer hidden states already provides a training-free, architecture-general feature basis. Individual dimensions encode semantic content via thei...
RODS: Reward-Driven Online Data Synthesis for Multi-Turn Tool-Use Agents
Multi-turn tool-use RL is bottlenecked by the rapid depletion of informative samples in static datasets. We observe that the gradient signal in GRPO concentrates on tasks with the ...
Deep Research in Physical Sciences: A Multi-Agent Framework and Comprehensive Benchmark
Deep research agents are Large Language Model (LLM)-based systems designed for autonomous, multi-step scientific reasoning, and they hold immense potential for accelerating researc...
ImageWAM: Do World Action Models Really Need Video Generation, or Just Image Editing?
World Action Models (WAMs) commonly rely on video generation to bridge visual world modeling and robot control. However, video-based WAMs face three coupled limitations: dense mult...
Moebius: 0.2B Lightweight Image Inpainting Framework with 10B-Level Performance
While 10B-level industrial foundation models have pushed the boundaries of image inpainting, their prohibitive computational costs severely hinder practical deployment. Constructin...
LooseControlVideo: Directorial Video Control using Spatial Blocking
Precise 3D spatial orchestration in text-to-video generation remains a significant challenge, particularly for multi-object scenes where semantic layout and temporal dynamics are o...
SproutRAG: Attention-Guided Tree Search with Progressive Embeddings for Long-Document RAG
Retrieval-augmented generation (RAG) systems must balance retrieval granularity with contextual coherence, a challenge that existing methods address through LLM-guided chunking, si...
Reinforcing Dual-Path Reasoning in Spatial Vision Language Models
Spatial VLMs have made substantial progress in geometric perception, yet complex spatial reasoning requiring multi-step inference over depth, distance, and scene relations remains ...
SAE Interventions are Unreliable: Post-Intervention Recovery of Suppressed Behavior
Sparse Autoencoders (SAEs) decompose residual-stream activations into interpretable features. Recent latent-space defenses increasingly rely on these decompositions, assuming that ...
Beyond NL2Code: A Structured Survey of Multimodal Code Intelligence
While Large Language Models (LLMs) have substantially advanced text-to-code synthesis, many real programming tasks specify intent through visual artifacts such as screenshots, char...
When Does Trajectory-Level Supervision Permit Efficient Offline Reinforcement Learning?
Offline reinforcement learning is typically analyzed under process-level reward supervision, yet many sequential decision datasets record only trajectory-level outcomes. We develop...
ChLogic: Evaluating Robustness of Logical Reasoning in Chinese Expressions
Large language models perform increasingly well on standardized logical reasoning benchmarks, but whether this ability remains robust beyond English is unclear. We introduce ChLogi...
Show the Signal, Hide the Noise: Spectral Forcing for Pixel-Space Diffusion
Pixel-space diffusion models are trained on full-bandwidth noisy images, yet the useful signal available to the denoiser is strongly frequency dependent. Under rectified-flow diffu...
Learning from the Self-future: On-policy Self-distillation for dLLMs
On-policy self-distillation (OPSD) has proven effective for post-training large language models (LLMs), yet its application to diffusion LLMs (dLLMs) remains unexplored. Existing O...
Adaptive Volumetric Mechanical Property Fields Invariant to Resolution
Accurate mechanical properties (or materials) Young's modulus (E), Poisson's ratio (ν) and density (ρ) are essential for reliable physics simulation of digital worlds, but most 3D ...
MCompassRAG: Topic Metadata as a Semantic Compass for Paragraph-Level Retrieval
Retrieval-augmented generation (RAG) systems depend critically on how documents are chunked and searched. Fine-grained chunks can improve retrieval precision but expand the search ...