awesome-long2short-on-lrms
Awesome-Long2short-on-LRMs is a collection of state-of-the-art, novel, exciting long2short methods on large reasoning models. It contains papers, codes, datasets, evaluations, and analyses.
https://github.com/hongcheng-gao/awesome-long2short-on-lrms
Last synced: 21 days ago
JSON representation
-
Length-Aware Guidance
-
Reward Guidance
-
link
- l3/l1) |
-
link
- Labs/efficient-reasoning) |
-
link
- Pruner) |
-
link
- tyx/puffin/main/recipe/dapo) |
-
link
- long-cot) |
-
Prompt Guidance
-
Others
-
Model Merge
-
link
-
link
- hkust/benchmark_inference_time_computation_LLM) |
-
link
- sg/LongSpec) |
-
Training Based
-
Parameter Modification
-
Computation Routing
-
Model Distillation
-
Routing Strategy
-
Model Merge
-
link
- to-short-via-model-merging) |
-
link
- |
-
Length-Agnostic Optimization
-
Routing Strategy
-
Latent Space Compression
-
Training Free
-
Prompt Guidance
-
Survey