Learns fine-grained preferences over sub-trajectories to identify and penalize redundant steps in long chain-of-thoughts, letting models "fold" reasoning chains into concise paths; reports ~56% token reduction on DeepSeek-R1-Distill-Qwen-7B while keeping accuracy.
Trains a GPT-style causal Transformer on a 2-billion-frame retargeted motion corpus to enable zero-shot whole-body motion tracking and control. By scaling both data and model capacity, it tracks highly dynamic behaviors while generalizing to unseen motions; accepted to CVPR 2026.
AI Model2026
Native multimodal model for image/text/video→text tasks with million‑token context support. Uses a sparse-attention operator to cut long‑context compute and latency, and targets agentic, coding, and long-horizon conversational workloads.
AI Model2026
Agentic LLM for long-horizon, environment-driven workflows: decomposes goals, generates and executes code/tool calls, evaluates outputs, and iterates. The Pro variant emphasizes coding and terminal execution and is published for use with sglang and multi-node H100 deployment.
AI Dataset2026
Generates synthetic coding-agent session traces by pairing remotely hosted open agent models with local llama.cpp user models across real open-source codebases. Each trace records read/write/edit/bash actions and tool use; the dataset is a reproducible cartesian product (20×3×20×20 = 24,000 sessions) under an MIT license.
Generates repository-specific LoRA adapters via a hypernetwork to inject repo-level knowledge into code LMs with zero inference-time token overhead. Provides a Static snapshot mode and an Evo mode that updates adapters per commit; evaluated on the 604-repo RepoPeftBench.
AI Model2026
Open-weight frontier LLM for agentic reasoning and long-context analysis (up to 1M tokens). Uses a LatentMoE + Mamba-2 hybrid with Multi-Token Prediction and NVFP4 efficiency (550B total / 55B active). Suited for multilingual agents, RAG, and heavy tool-use workloads.
AI Audio2026
Converts text into expressive conversational speech across 100+ languages with zero-shot voice cloning and inline control tokens for emotion, style, prosody, pauses, and sound effects. Released under a research/non-commercial license; commercial use requires separate licensing.
AI Model2026
Multilingual frontier LLM optimized for long-context reasoning and agentic workflows, combining a LatentMoE (Mamba-2 + MoE) hybrid architecture with Multi-Token Prediction and NVFP4 quantization; targeted for NVIDIA GPU deployments and governed by the OpenMDW-1.1 license.
AI Model2026
Provides compact, agentic text-generation for long-horizon, tool-enabled workflows — trading some peak capability for lower latency and easier on-prem deployment. Key features: adaptive/coherent thinking traces, function-calling support, and sglang/docker-ready serving.
AI Model2026
Provides a GGUF-ready QAT (Q4_0) quantized build of Gemma 4 12B that preserves near-bfloat16 quality while reducing memory footprint for local inference; compatible with Transformers-based and GGUF runtimes.
AI Model2026
A surgically modified Gemma 4 (12B) that removes refusal behavior while preserving benchmark parity; released as an uncensored research artifact with GGUF quantizations for local inference and red‑team/alignment evaluation.