CRISP-Qwen3-8B-v1

Qwen3-8B trained with CRISP (Compressed Reasoning via Iterative Self-Policy Distillation) using the v1 conciseness teacher. Step-99 checkpoint.

Paper: https://arxiv.org/abs/2603.05433

CRISP teaches a reasoning model to think concisely by distilling its own concise behavior back into itself: the teacher is the same model conditioned on a conciseness instruction, the student has no instruction, and training minimizes per-token reverse KL from student to teacher on the student's own rollouts (teacher refreshed every M=50 steps). No ground-truth answers, token budgets, or difficulty estimators enter the loss.

This checkpoint uses the v1 teacher prompt: v1 (uniform): "Solve concisely and correctly. Be direct — avoid unnecessary elaboration, redundant steps, or restating the problem."

Other CRISP checkpoints: Qwen3-8B (v2), Qwen3-14B (v2), DeepSeek-R1-Distill-Llama-8B (v2). Training data: pb09204048/CRISP.

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer
tok = AutoTokenizer.from_pretrained("pb09204048/CRISP-Qwen3-8B-v1")
model = AutoModelForCausalLM.from_pretrained("pb09204048/CRISP-Qwen3-8B-v1", device_map="auto")

Benchmark results (Qwen3-8B)

Accuracy (mean@8, %) and token reduction (Red., % vs. base) at a 30K-token budget. Math is scored with a dual-path grader (Answer: or \boxed{}); GPQA-Diamond and MMLU use exact letter-match. This model is the CRISP (v1) row.

Citation

@article{sang2026crisp,
  title={Crisp: Compressed reasoning via iterative self-policy distillation},
  author={Sang, Hejian and Xu, Yuanda and Zhou, Zhengze and He, Ran and Wang, Zhipeng and Sun, Jiachen},
  journal={arXiv preprint arXiv:2603.05433},
  year={2026}
}