Trust or Escalate: LLM Judges with Provable Guarantees for Human Agreement

• URL: http://arxiv.org/abs/2407.18370v1
• Date: Thu, 25 Jul 2024 20:04:59 GMT
• Title: Trust or Escalate: LLM Judges with Provable Guarantees for Human Agreement
• Authors: Jaehun Jung, Faeze Brahman, Yejin Choi,
• Abstract summary: We present a principled approach to provide LLM-based evaluation with a rigorous guarantee of human agreement. We first propose that a reliable evaluation method should not uncritically rely on model preferences for pairwise evaluation. We then show that under this selective evaluation framework, human agreement can be provably guaranteed.
• Score: 49.15348173246146
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We present a principled approach to provide LLM-based evaluation with a rigorous guarantee of human agreement. We first propose that a reliable evaluation method should not uncritically rely on model preferences for pairwise evaluation, but rather assess the confidence of judge models and selectively decide when to trust its judgement. We then show that under this selective evaluation framework, human agreement can be provably guaranteed -- such that the model evaluation aligns with that of humans to a user-specified agreement level. As part of our framework, we also introduce Simulated Annotators, a novel confidence estimation method that significantly improves judge calibration and thus enables high coverage of evaluated instances. Finally, we propose Cascaded Selective Evaluation, where we use cheaper models as initial judges and escalate to stronger models only when necessary -- again, while still providing a provable guarantee of human agreement. Experimental results show that Cascaded Selective Evaluation guarantees strong alignment with humans, far beyond what LLM judges could achieve without selective evaluation. For example, on a subset of Chatbot Arena where GPT-4 almost never achieves 80% human agreement, our method, even while employing substantially cost-effective models such as Mistral-7B, guarantees over 80% human agreement with almost 80% test coverage.

Related papers

• CompassJudger-1: All-in-one Judge Model Helps Model Evaluation and Evolution [74.41064280094064]
  textbfJudger-1 is the first open-source textbfall-in-one judge LLM. CompassJudger-1 is a general-purpose LLM that demonstrates remarkable versatility. textbfJudgerBench is a new benchmark that encompasses various subjective evaluation tasks.
  arXiv  Detail & Related papers  (2024-10-21T17:56:51Z)
• Limits to scalable evaluation at the frontier: LLM as Judge won't beat twice the data [14.95829896035971]
  An emerging family of debiasing tools promises to fix issues by using a few high quality labels to debias a large number of model judgments. Our main result shows that when the judge is no more accurate than the evaluated model, no debiasing method can decrease the required amount of ground truth labels by more than half.
  arXiv  Detail & Related papers  (2024-10-17T08:49:42Z)
• Self-rationalization improves LLM as a fine-grained judge [21.917301609125417]
  We introduce Self-Rationalization, an iterative process of improving the rationales for the judge models. Self-rationalization works by having the model generate multiple judgments with rationales for the same input. We show that our model learns to produce higher quality rationales, with a win rate of $62%$ on average compared to models just trained via SFT on rationale.
  arXiv  Detail & Related papers  (2024-10-07T21:05:53Z)
• Poor-Supervised Evaluation for SuperLLM via Mutual Consistency [20.138831477848615]
  We propose the PoEM framework to conduct evaluation without accurate labels. We first prove that the capability of a model can be equivalently assessed by the consistency between it and certain reference model. To alleviate the insufficiencies of the conditions in reality, we introduce an algorithm that treats humans (when available) and the models under evaluation as reference models.
  arXiv  Detail & Related papers  (2024-08-25T06:49:03Z)
• Judging the Judges: Evaluating Alignment and Vulnerabilities in LLMs-as-Judges [6.609843448260634]
  The LLM-as-a-judge paradigm is rapidly gaining traction as an approach to evaluating large language models. This paper focuses on a clean scenario in which inter-human agreement is high. We identify vulnerabilities in judge models, such as their sensitivity to prompt complexity and length, and a tendency toward leniency.
  arXiv  Detail & Related papers  (2024-06-18T13:49:54Z)
• From Adversarial Arms Race to Model-centric Evaluation: Motivating a Unified Automatic Robustness Evaluation Framework [91.94389491920309]
  Textual adversarial attacks can discover models' weaknesses by adding semantic-preserved but misleading perturbations to the inputs. The existing practice of robustness evaluation may exhibit issues of incomprehensive evaluation, impractical evaluation protocol, and invalid adversarial samples. We set up a unified automatic robustness evaluation framework, shifting towards model-centric evaluation to exploit the advantages of adversarial attacks.
  arXiv  Detail & Related papers  (2023-05-29T14:55:20Z)
• Large Language Models are not Fair Evaluators [60.27164804083752]
  We find that the quality ranking of candidate responses can be easily hacked by altering their order of appearance in the context. This manipulation allows us to skew the evaluation result, making one model appear considerably superior to the other. We propose a framework with three simple yet effective strategies to mitigate this issue.
  arXiv  Detail & Related papers  (2023-05-29T07:41:03Z)
• Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback [91.22679548111127]
  A trustworthy real-world prediction system should produce well-calibrated confidence scores. We show that verbalized confidences emitted as output tokens are typically better-calibrated than the model's conditional probabilities.
  arXiv  Detail & Related papers  (2023-05-24T10:12:33Z)
• Investigating Crowdsourcing Protocols for Evaluating the Factual Consistency of Summaries [59.27273928454995]
  Current pre-trained models applied to summarization are prone to factual inconsistencies which misrepresent the source text or introduce extraneous information. We create a crowdsourcing evaluation framework for factual consistency using the rating-based Likert scale and ranking-based Best-Worst Scaling protocols. We find that ranking-based protocols offer a more reliable measure of summary quality across datasets, while the reliability of Likert ratings depends on the target dataset and the evaluation design.
  arXiv  Detail & Related papers  (2021-09-19T19:05:00Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.