CLAISDMay 26

GrowLoop: Self-Evolving Conversation Evaluation Seeded by Human

arXiv:2605.2888291.6h-index: 1
Predicted impact top 26% in CL · last 90 daysOriginality Highly original
AI Analysis

For researchers evaluating human-likeness in LLM conversations, GrowLoop addresses the challenges of implicit criteria, human disagreement, and evolving standards with a continuously adaptive system.

GrowLoop introduces a self-evolving evaluation system for open-ended conversation that continuously adapts to advancing models and shifting scenarios, achieving substantially better alignment with human judgments than existing methods and effectively discriminating model capability tiers.

With the rapid advancement of large language models, evaluating human-likeness in open-ended conversation has become increasingly important. However, human-likeness is a form of tacit knowledge that humans perceive intuitively, yet the underlying criteria resist explicit formulation. Human judgments vary widely, with strong agreement on some cases and legitimate disagreement on others. Meanwhile, the criteria behind human judgments remain implicit, leaving no clear basis for constructing cases. Further, what counts as human-like is not static, but evolving with model capability and human expectations. Despite progress in evaluation methods such as expert-authored benchmarks, Reward Models, and self-evolving benchmarks, none addresses all three challenges simultaneously. Therefore, we propose GrowLoop, a self-evolving conversation evaluation system that continuously adapts as models advance and scenarios shift. With minimal human seed annotations as the first mover, LLM agents iteratively extract and refine evaluation rubrics through Heuristic Learning. Human-AI agreement is required where annotators converge, while only plausibility is expected where they diverge. Moreover, the Rubric-Case co-evolution mechanism enables continuous evolution, expanded through new seeds when the evaluation target moves. Applied to human-likeness evaluation in open-ended conversation, the generated rubrics not only substantially outperform existing methods in alignment with human judgments, but also uncover issues that annotators overlook. The resulting benchmark effectively discriminates models across capability tiers and reveals where they fall short, while generalizing to new scenarios and adapting as models advance. Our work shifts the benchmarking paradigm from manual updates or difficulty scaling to comprehensive, continuous self-evolution.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes