sail-sg / LightTransLinks
The official implementation of "LightTransfer: Your Long-Context LLM is Secretly a Hybrid Model with Effortless Adaptation"
☆22Updated 9 months ago
Alternatives and similar repositories for LightTrans
Users that are interested in LightTrans are comparing it to the libraries listed below
Sorting:
- CoT-Valve: Length-Compressible Chain-of-Thought Tuning☆91Updated 11 months ago
- ☆20Updated 9 months ago
- [ICLR 2026] dParallel: Learnable Parallel Decoding for dLLMs☆58Updated 2 weeks ago
- ☆46Updated 4 months ago
- [ICML 2024] Unveiling and Harnessing Hidden Attention Sinks: Enhancing Large Language Models without Training through Attention Calibrati…☆46Updated last year
- Official Repository of LatentSeek☆76Updated 8 months ago
- [ICLR2026] Laser: Learn to Reason Efficiently with Adaptive Length-based Reward Shaping☆62Updated 8 months ago
- [ICML'25] Our study systematically investigates massive values in LLMs' attention mechanisms. First, we observe massive values are concen…☆85Updated 7 months ago
- [COLM 2025] SEAL: Steerable Reasoning Calibration of Large Language Models for Free☆51Updated 10 months ago
- A Sober Look at Language Model Reasoning☆92Updated 2 months ago
- ☆74Updated 9 months ago
- The loss landscape of Large Language Models resemble basin!☆36Updated 7 months ago
- Model merging is a highly efficient approach for long-to-short reasoning.☆98Updated 3 months ago
- [ICML'25] Official code of paper "Fast Large Language Model Collaborative Decoding via Speculation"☆28Updated 7 months ago
- ☆63Updated 6 months ago
- Codebase for decoding compressed trust.☆25Updated last year
- ☆23Updated last year
- ☆145Updated 4 months ago
- Official repository for paper: O1-Pruner: Length-Harmonizing Fine-Tuning for O1-Like Reasoning Pruning☆97Updated 11 months ago
- The official github repo for "Training Optimal Large Diffusion Language Models", the first-ever large-scale diffusion language models sca…☆45Updated 3 months ago
- Official code for Guiding Language Model Math Reasoning with Planning Tokens☆18Updated last year
- Extending context length of visual language models☆12Updated last year
- ☆47Updated 10 months ago
- [ICML 2025] M-STAR (Multimodal Self-Evolving TrAining for Reasoning) Project. Diving into Self-Evolving Training for Multimodal Reasoning☆70Updated 6 months ago
- [NeurIPS 2025] Think Silently, Think Fast: Dynamic Latent Compression of LLM Reasoning Chains☆74Updated 6 months ago
- [ICLR 2025] When Attention Sink Emerges in Language Models: An Empirical View (Spotlight)☆154Updated 7 months ago
- AdaRFT: Efficient Reinforcement Finetuning via Adaptive Curriculum Learning☆54Updated 7 months ago
- [EMNLP 2025] TokenSkip: Controllable Chain-of-Thought Compression in LLMs☆201Updated 2 months ago
- RL with Experience Replay☆55Updated 6 months ago
- Optimizing Anytime Reasoning via Budget Relative Policy Optimization☆51Updated 6 months ago