[KongchangAI]
Model

JurisCoT-SFT

The first-stage supervised fine-tuning method in a two-stage training framework for legal reasoning large models, which balances supervision signals for reasoning chains and answers through length-normalized dual-objective learning.

Timeline (last 90 days)

Nov 13

The paper proposes JurisCoT-SFT, which computes losses separately for the reasoning chain and the final answer and normalizes them by sequence length so that the learning signals of the two parts are on a comparable scale.

Unverified50%
Nov 13

Experimental results show that the framework achieves stable improvements in both reasoning chain quality and final answer accuracy, and the model remains competitive when compared with larger-scale models.

Unverified50%
Nov 13

The paper emphasizes that the improvement is in 'consistency'—not a surge in some single metric, but simultaneous improvement in both reasoning chain quality and final answer accuracy.

Unverified50%
Nov 13

The generality of the framework still needs verification, and the degree to which the two-stage training depends on data quality, as well as the stability of the reward scheduling strategy across different legal systems, are issues worthy of attention in follow-up work.

Unverified50%

All Facts (4)

Source Articles