Evaluates multimodal LLMs' ability to reconstruct past observations and act in controllable non-Markov games. Introduces RNG-Bench with two games (Matching Pairs, 3D Maze), three controllable difficulty axes, a head-to-head duel protocol, and a Memory Gap metric to separate forgetting from action errors.
Generates images from natural-language prompts as an 8-step distilled checkpoint of Krea 2, optimized for fast iterative text-to-image workflows with style references and 1K–2K resolution outputs.
Provides GGUF-quantized weights and runtime assets for running the Qwythos-9B reasoning LLM locally via llama.cpp and compatible runtimes. Key features include 1,048,576-token YaRN long-context, native function-calling, multimodal image input (requires mmproj), and multiple quantization/MTP variants tuned for different size/quality tradeoffs.
A 9B reasoning LLM fine-tuned from Qwen3.5 that ships with a 1,048,576-token context, native function-calling and tool-use, and notable benchmark gains (+34 MMLU, +30 gsm8k-strict).
A dense ~9B reasoning LLM optimized for agentic coding and tool-calling that emits explicit chain-of-thought (<think>) blocks and well-formed tool calls. Designed to run on a single 80GB GPU (~19GB bf16), uses self-scaffolding RL and exposes an OpenAI-compatible API.
Generates streaming, low‑latency neural speech for real‑time dialogue by autoregressively producing audio frames as text arrives; joint text–speech training preserves natural prosody. Optimized for vLLM streaming (~50 ms first chunk), supports short‑clip voice cloning and four languages.
Simulates agentic environments and predicts next environment states from actions and interaction history using a language-based world model across seven domains. Trained via a CPT→SFT→RL pipeline with an MoE architecture and very long context; intended for environment simulation and agent research.
Provides 2,170 reference-grounded evaluation samples across seven agent domains (MCP, Search, Terminal, SWE, Android, Web, OS) to score language world models on Format, Factuality, Consistency, Realism and Quality. Includes per-domain JSONL files, judge prompts and an evaluation script for reproducible scoring.
NVFP4-quantized variant of Qwen3.6-27B that reduces parameter bits from 16 to 4, cutting disk and GPU memory requirements by ~2.5× while keeping comparable benchmark accuracy; ready for vLLM-based inference on NVIDIA hardware and supports long, multimodal contexts.
Compresses an entire document collection into 16 query-dependent memory tokens for a frozen LLM. It retrieves useful clues across long histories while keeping the model’s input length fixed.
Provides an open-source Mixture-of-Experts coding LLM (397B) optimized for agentic, tool-enabled coding workflows with a 262,144-token context window, OpenAI-compatible API, serving recipes (vLLM/SGLang), and published coding-benchmark results.
Proposes Monotonic Inference Policy Improvement (MIPI) and a two-step Monotonic Inference Policy Update (MIPU) to address training–inference probability mismatch in LLM reinforcement learning by constructing sampler-referenced candidate updates and accepting synchronized updates using an inference-gap proxy; shows improved reasoning accuracy and stability under FP8-quantized rollouts.