Fetching the paper…

TL;DR: Too Long, Do Re-weighting for Efficient LLM Reasoning Compression · Around