Provides 7,366 recorded agent trajectories from H Company’s Holo4 benchmark runs, with step-level reasoning, actions, tool results, token usage and screenshots for replay and analysis. Bundled as JSON and image files for per-trajectory inspection and automated replay; released under Apache 2.0.
Teaches small reasoning models to diagnose when additional internal thinking is insufficient and to selectively query stronger models; introduces FlyBy, a supervised + cost-aware RL framework that learns when/what to ask. Improves pass@ metrics on hard benchmarks while reducing serving cost.
Improves test-time scaling of looped transformers by adaptively assigning extra recurrent iterations to tokens that benefit most. TaH2 is a post-training method that jointly trains an iteration decider with the backbone using lookahead depth supervision, boosting the accuracy–compute slope and peak accuracy on challenging benchmarks.
Trains a single unified multimodal model with reinforcement learning to perform end-to-end self-reflection and iterative image repair — jointly learning the diagnostic (textual) reflection and the flow-based image revisions so credit propagates across rounds without an external verifier.
Defines and evaluates AREX-2, an LLM agent that iteratively self-improves at test time via reflection and long-horizon execution. Trained on long-horizon improvement trajectories from ML engineering and algorithmic programming (built on Qwen3.8-27B), it scales with more rounds and achieves strong benchmark scores.
Lets general-purpose vision-language models directly command robots via a compact mid-level action interface and asynchronous monitoring, enabling zero-shot manipulation without task-specific policy training; demonstrates strong sim benchmarks and real xArm6 transfer.
Groups visually grounded appearances of the same physical instance into persistent, retrievable “biographies” so agents can follow objects across hours or days for long-video question answering. Links identity-aware observations to episodic context and visual evidence; improves EgoLifeQA to 72.0% and increases evidence-window reach from 37.6% to 58.9%.
Turns solved protein structures into FoldingCorpus and Fold2Reason — a post-training recipe that supervises an LLM with discrete structural Q&A plus continuous 3D geometry to improve spatial, graph and scientific reasoning; reports consistent gains across 10 benchmarks.
Delivers fast, calibrated 'System 1' decisions over text and images (yes/no, 2–256 choices, 0–5 scores) in one forward pass, plus full 'System 2' Qwen3.8-27B multimodal reasoning. Highlights: zero-shot vision judging, 256K-token context, vLLM decision API, Apache-2.0.
On-device speech-to-text for short clips (up to 30s) in seven languages, yielding transcripts, word-level timestamps and per-frame speech embeddings in a single 16.9 MB model. Runs on the Needle CPU engine with 2–4 bit quantization, supports keyword biasing and returns empty transcripts for silence.
Turns a structured state (text/JSON/images/video) plus a schema of typed questions into probabilistic decisions in one forward pass. Multimodal, Jev/SystemOne-compatible, 64k context window, post-trained from Qwen3.8-27B; no free-form text output.
Assesses how AI reviewers react to content-preserving rewrites and introduces RobustReview, a controlled benchmark plus SciCore, a dual-branch reviewer that combines full-manuscript judgment with a structured 'science core' summary to reduce rhetorical sensitivity.