google-deepmind / latent-multi-hop-reasoningLinks
[ACL 2024] Do Large Language Models Latently Perform Multi-Hop Reasoning?
☆79Updated 7 months ago
Alternatives and similar repositories for latent-multi-hop-reasoning
Users that are interested in latent-multi-hop-reasoning are comparing it to the libraries listed below
Sorting:
- OpenCoconut implements a latent reasoning paradigm where we generate thoughts before decoding.☆172Updated 9 months ago
- [EMNLP 2025] The official implementation for paper "Agentic-R1: Distilled Dual-Strategy Reasoning"☆101Updated last month
- ☆135Updated 7 months ago
- [ACL 2025] Agentic Reward Modeling: Integrating Human Preferences with Verifiable Correctness Signals for Reliable Reward Systems☆108Updated 4 months ago
- ☆68Updated 5 months ago
- Entropy Based Sampling and Parallel CoT Decoding☆17Updated last year
- ☆58Updated 4 months ago
- Training an LLM to use a calculator with multi-turn reinforcement learning, achieving a **62% absolute increase in evaluation accuracy**.☆57Updated 5 months ago
- ☆35Updated 5 months ago
- Train your own SOTA deductive reasoning model☆108Updated 7 months ago
- Simple & Scalable Pretraining for Neural Architecture Research☆297Updated 2 months ago
- Archon provides a modular framework for combining different inference-time techniques and LMs with just a JSON config file.☆187Updated 7 months ago
- Official PyTorch implementation for Hogwild! Inference: Parallel LLM Generation with a Concurrent Attention Cache☆127Updated 2 months ago
- ☆122Updated 8 months ago
- Source code for the collaborative reasoner research project at Meta FAIR.☆102Updated 6 months ago
- Storing long contexts in tiny caches with self-study☆201Updated last week
- Super basic implementation (gist-like) of RLMs with REPL environments.☆204Updated last week
- rl from zero pretrain, can it be done? yes.☆277Updated 3 weeks ago
- Optimizing Causal LMs through GRPO with weighted reward functions and automated hyperparameter tuning using Optuna☆58Updated last week
- A fast, local, and secure approach for training LLMs for coding tasks using GRPO with WebAssembly and interpreter feedback.☆39Updated 6 months ago
- Matrix (Multi-Agent daTa geneRation Infra and eXperimentation framework) is a versatile engine for multi-agent conversational data genera…☆97Updated this week
- ☆218Updated 8 months ago
- ☆55Updated 11 months ago
- accompanying material for sleep-time compute paper☆117Updated 5 months ago
- ☆27Updated 4 months ago
- j1-micro (1.7B) & j1-nano (600M) are absurdly tiny but mighty reward models.☆98Updated 3 months ago
- Analysis code for Neurips 2025 paper "SciArena: An Open Evaluation Platform for Foundation Models in Scientific Literature Tasks"☆53Updated 2 months ago
- QAlign is a new test-time alignment approach that improves language model performance by using Markov chain Monte Carlo methods.☆24Updated last week
- Code for our paper PAPILLON: PrivAcy Preservation from Internet-based and Local Language MOdel ENsembles☆59Updated 5 months ago
- EvaByte: Efficient Byte-level Language Models at Scale☆110Updated 6 months ago