📖 This is a repository for organizing papers, codes, and other resources related to Latent Reasoning.

Speed Always Wins: A Survey on Efficient Architectures for Large Language Models

paper list, tutorial, and nano code snippet for Diffusion Large Language Models.

Experimental playground for benchmarking language model (LM) architectures, layers, and tricks on smaller datasets. Designed for flexible experimentation and exploration.

[ICLR 2026] ParallelBench: Understanding the Tradeoffs of Parallel Decoding in Diffusion LLMs

Educational Implementation of "Edit Flows: Flow Matching with Edit Operations" by Havasi et al.

Anchored Diffusion Language Model (NeurIPS 2025)

A Curated Collection of Frontier Language Model Architectures

[ACL 2026 Main] Code for paper "d-TreeRPO: Towards More Reliable Policy Optimization for Diffusion Language Models"

[ICLR2026] Official Code for Loopholing Discrete Diffusion: Deterministic Bypass of the Sampling Wall

Quantization Meets dLLMs: A Systematic Study of Post-training Quantization for Diffusion LLMs

Auto-setup Inception Labs Mercury-2 to 9router

FastMCP fleet MCP server for diffusion LMs (dLLM). DiffusionGemma on Goliath RTX 4090 — batch inference, HLE-shaped reasoning, ~200–400 tok/s. Doc phase; llama-diffusion-cli sidecar next. Complements local-llm-mcp.

A curated list of diffusion language models (dLLMs): papers, models, and code behind diffusion-based, non-autoregressive text generation, focused on 2025–2026.

Diffusion VLM papers indexed with one feature schema, plus decode-behaviour traces measured through a single instrumentation path

An Apple Silicon / MLX research lab exploring AR-to-diffusion conversion of Qwen3.5, including failed bidirectional Gated DeltaNet experiments and a FLARE-inspired reproduction.

Trying out Diffusion LLMS