Token warping — rearranging ViT image tokens rather than pixels — enables MLLMs to reason from nearby viewpoints without fine-tuning, consistently outperforming all baselines on the new ViewBench benchmark.
Current top AI models score just 56.3% on tasks humans ace at 93.8% — this benchmark finally exposes exactly where and why multimodal agents fail.
RLSD solves the information leakage problem of on-policy self-distillation by repurposing the teacher as a token-level magnitude evaluator, achieving state-of-the-art on 5 multimodal reasoning benchmarks.
A minimal 4-frame sliding-window baseline beats all published streaming video models with half the GPU memory.
A coding agent that automatically discovers better LLM harnesses—achieving rank #1 on TerminalBench-2 and +7.7 points over ACE on text classification, using filesystem access for causal diagnosis.
DataFlex unifies dynamic data selection, domain mixture optimization, and sample reweighting into a single LLaMA-Factory-compatible framework, enabling reproducible data-centric LLM training with consistent MMLU gains.
PixelSmile achieves precise facial expression editing at the pixel level, overcoming semantic overlap issues that plagued prior methods.
Voxtral TTS by Mistral AI generates highly natural multilingual speech from minimal data, setting a new standard for expressive text-to-speech.
RealRestorer leverages large-scale image editing models to achieve generalizable real-world image restoration, tackling complex degradations that previous methods couldn't handle.
Intern-S1-Pro is the world's first trillion-parameter scientific multimodal model, setting new state-of-the-art across 50+ scientific benchmarks.
This paper introduces Natural-Language Agent Harnesses (NLAHs), showing that agent control logic can be expressed in editable text rather than code — achieving a 55% performance boost when migrating from code to natural language.
Calibri reveals hidden capacity in Diffusion Transformers through lightweight parameter-efficient calibration, achieving significant quality gains with minimal compute overhead.
A three-layer security framework for autonomous AI agents that protects against data leakage, privilege escalation, and malicious tool execution in real-time.
CUA-Suite unlocks computer-use agent training at scale with 55 hours of expert-annotated desktop video — the largest benchmark of its kind.
A simple architectural modification to Transformers that feeds attention outputs back as residuals, improving reasoning and long-context performance without additional parameters.
Could AI achieve superintelligence by building its own tools? This paper argues yes — through the Diligent Learner framework combining test-time search with tool-building.
AutoHarness teaches a smaller LLM to write a code guardrail that blocks illegal moves — letting Gemini-2.5-Flash outperform Gemini-2.5-Pro across 145 games.
PaperBanana automates publication-ready academic illustrations using VLM-powered agents — a potential game-changer for the AI research workflow.
A unified foundation model that jointly generates synchronized audio and video from text prompts, eliminating the need for separate audio and video pipelines.
ACE (Agentic Context Engineering) treats LLM contexts as evolving playbooks, achieving +10.6% on agent benchmarks through systematic context optimization.
VibeVoice synthesizes 90-minute, 4-speaker conversations using next-token diffusion with a 7.5 Hz tokenizer that compresses speech 80× vs Encodec — making long-form multi-speaker TTS feasible in a standard LLM context window.
LlamaFactory provides a unified framework for fine-tuning 100+ language models with minimal code, supporting LoRA, QLoRA, RLHF and more out of the box.
CARLA-Air unifies drone flight and autonomous driving in a single simulation, enabling air-ground cooperative AI research without co-simulation overhead.
InCoder-32B-Thinking bridges general and industrial code intelligence through Error-driven Chain-of-Thought synthesis and an Industrial Code World Model, achieving 81.3% on LiveCodeBench and 84.0% on CAD-Coder.
