[KongchangAI]
Model

LegalEvolver

A query optimization framework for legal judgment prediction (LJP). Through heuristic seed queries, rubric-guided multi-round self-evolution, and reinforcement learning, it generates retrieval queries rich in legal information to improve the performance of legal AI agents.

Timeline (last 90 days)

Nov 5

LegalEvolver is a query optimization framework specifically designed to generate legally information-rich queries for LJP-oriented retrieval.

Unverified50%
Nov 5

Directly sampling high-quality optimized queries and their corresponding search trajectories is very difficult; relying solely on unguided high-temperature random sampling is not only inefficient but also limited by the model's initial capabilities.

Unverified50%
Nov 5

LegalEvolver first optimizes queries through heuristic methods, uses them as seeds, and then generates search trajectories based on these seeds.

Unverified50%
Nov 5

LegalEvolver summarizes the experience of each rollout into scoring rubrics and uses these rubrics to guide a multi-round iterative self-evolution process; the model continuously optimizes the scoring rubrics and uses the optimized rubrics to guide reinforcement learning (RL).

Unverified50%
Nov 5

Comparative experiments in the paper show that evolving scoring rubrics are more effective than fixed rubrics.

Unverified50%
Nov 5

Compared with the approach of directly applying supervised fine-tuning (SFT) followed by reinforcement learning, the rubric-guided method performs better.

Unverified50%
Nov 5

In the context of this paper, scoring rubrics are not a manually pre-fixed scoring table, but rather a set of natural-language criteria that the model summarizes from sampling experience and uses to judge the quality of queries.

Unverified50%
Nov 5

The self-evolution mechanism enables the system to surpass the performance ceiling set by the initial rubrics or the initial model capabilities.

Unverified50%
Nov 5

The paper leaves open questions: how to guarantee the quality of the scoring rubrics, and whether the self-evolution process introduces bias.

Unverified50%

All Facts (9)

Unverified

LegalEvolver is a query optimization framework specifically designed to generate legally information-rich queries for LJP-oriented retrieval.

50%
Unverified

Directly sampling high-quality optimized queries and their corresponding search trajectories is very difficult; relying solely on unguided high-temperature random sampling is not only inefficient but also limited by the model's initial capabilities.

50%
Unverified

LegalEvolver first optimizes queries through heuristic methods, uses them as seeds, and then generates search trajectories based on these seeds.

50%
Unverified

LegalEvolver summarizes the experience of each rollout into scoring rubrics and uses these rubrics to guide a multi-round iterative self-evolution process; the model continuously optimizes the scoring rubrics and uses the optimized rubrics to guide reinforcement learning (RL).

50%
Unverified

Comparative experiments in the paper show that evolving scoring rubrics are more effective than fixed rubrics.

50%
Unverified

Compared with the approach of directly applying supervised fine-tuning (SFT) followed by reinforcement learning, the rubric-guided method performs better.

50%
Unverified

In the context of this paper, scoring rubrics are not a manually pre-fixed scoring table, but rather a set of natural-language criteria that the model summarizes from sampling experience and uses to judge the quality of queries.

50%
Unverified

The self-evolution mechanism enables the system to surpass the performance ceiling set by the initial rubrics or the initial model capabilities.

50%
Unverified

The paper leaves open questions: how to guarantee the quality of the scoring rubrics, and whether the self-evolution process introduces bias.

50%

Source Articles