CLM-8B: Contrastive Language Model Beats DeepSWE at 9x Speed

A team released CLM-8B, a contrastive language model connecting states and actions via contrastive learning. With light fine-tuning it surpasses DeepSWE at 81.6% while running inference up to 9x faster.

2026-09-24 ~ 2026-09-24 · 2 related posts

• CLM-8B: contrastive System One model claims 9x faster inference, agentic SOTA — ChengleiSi · 2026-09-24
• CLM-8B hits SOTA 81.6% on DeepSWE with light finetuning, up to 9x faster inference — anshulkundaje · 2026-09-24