RankAlign: A Ranking View of the Generator-Validator Gap in Large Language Models

• URL: http://arxiv.org/abs/2504.11381v1
• Date: Tue, 15 Apr 2025 16:53:31 GMT
• Title: RankAlign: A Ranking View of the Generator-Validator Gap in Large Language Models
• Authors: Juan Diego Rodriguez, Wenxuan Ding, Katrin Erk, Greg Durrett,
• Abstract summary: We consider the discrepancy between a model's generated answer and their own verification of that answer, the generator-validator gap.<n>We show that according to this measure, a large gap exists in various settings, including question answering, lexical semantics tasks, and next-word prediction.<n>We then propose RankAlign, a ranking-based training method, and show that it significantly closes the gap by 31.8% on average, surpassing all baseline methods.
• Score: 51.080608392304505
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Although large language models (LLMs) have become generally more capable and accurate across many tasks, some fundamental sources of unreliability remain in their behavior. One key limitation is their inconsistency at reporting the the same information when prompts are changed. In this paper, we consider the discrepancy between a model's generated answer and their own verification of that answer, the generator-validator gap. We define this gap in a more stringent way than prior work: we expect correlation of scores from a generator and a validator over the entire set of candidate answers. We show that according to this measure, a large gap exists in various settings, including question answering, lexical semantics tasks, and next-word prediction. We then propose RankAlign, a ranking-based training method, and show that it significantly closes the gap by 31.8% on average, surpassing all baseline methods. Moreover, this approach generalizes well to out-of-domain tasks and lexical items.

Related papers

• GRAPE: Let GPRO Supervise Query Rewriting by Ranking for Retrieval [19.73916326078242]
  The CLIP model has become a cornerstone of large-scale retrieval systems by aligning text and image data in a unified embedding space.<n>To avoid costly retraining, existing methods mainly adopt query-rewriting strategies with large language models (LLMs)<n>We address this challenge with GRAPE, a plug-and-play enhancement approach that incorporates ranking signals into retrieval-guided query rewriting.
  arXiv  Detail & Related papers  (2025-09-27T15:36:59Z)
• Threading the Needle: Reweaving Chain-of-Thought Reasoning to Explain Human Label Variation [44.25455164977285]
  Large Language Models (LLMs) generate chains of thought (CoTs) before giving the final answer.<n>We propose a novel pipeline enriched with linguistically-grounded discourse segmenters to extract supporting and opposing statements for each answer option.<n>We also propose a rank-based HLV evaluation framework that prioritizes the ranking of answers over exact scores.
  arXiv  Detail & Related papers  (2025-05-29T11:47:18Z)
• When Disagreements Elicit Robustness: Investigating Self-Repair Capabilities under LLM Multi-Agent Disagreements [56.29265568399648]
  We argue that disagreements prevent premature consensus and expand the explored solution space.<n>Disagreements on task-critical steps can derail collaboration depending on the topology of solution paths.
  arXiv  Detail & Related papers  (2025-02-21T02:24:43Z)
• Pointwise Mutual Information as a Performance Gauge for Retrieval-Augmented Generation [78.28197013467157]
  We show that the pointwise mutual information between a context and a question is an effective gauge for language model performance.<n>We propose two methods that use the pointwise mutual information between a document and a question as a gauge for selecting and constructing prompts that lead to better performance.
  arXiv  Detail & Related papers  (2024-11-12T13:14:09Z)
• Generative Multi-modal Models are Good Class-Incremental Learners [51.5648732517187]
  We propose a novel generative multi-modal model (GMM) framework for class-incremental learning. Our approach directly generates labels for images using an adapted generative model. Under the Few-shot CIL setting, we have improved by at least 14% accuracy over all the current state-of-the-art methods with significantly less forgetting.
  arXiv  Detail & Related papers  (2024-03-27T09:21:07Z)
• ReAGent: A Model-agnostic Feature Attribution Method for Generative Language Models [4.015810081063028]
  Feature attribution methods (FAs) are employed to derive the importance of all input features to the model predictions. It is unknown if it is faithful to use these FAs for decoder-only models on text generation. We present a model-agnostic FA for generative LMs called Recursive Attribution Generator (ReAGent)
  arXiv  Detail & Related papers  (2024-02-01T17:25:51Z)
• See, Say, and Segment: Teaching LMMs to Overcome False Premises [67.36381001664635]
  We propose a cascading and joint training approach for LMMs to solve this task. Our resulting model can "see" by detecting whether objects are present in an image, "say" by telling the user if they are not, and finally "segment" by outputting the mask of the desired objects if they exist.
  arXiv  Detail & Related papers  (2023-12-13T18:58:04Z)
• Generative Judge for Evaluating Alignment [84.09815387884753]
  We propose a generative judge with 13B parameters, Auto-J, designed to address these challenges. Our model is trained on user queries and LLM-generated responses under massive real-world scenarios. Experimentally, Auto-J outperforms a series of strong competitors, including both open-source and closed-source models.
  arXiv  Detail & Related papers  (2023-10-09T07:27:15Z)
• Mitigating Word Bias in Zero-shot Prompt-based Classifiers [55.60306377044225]
  We show that matching class priors correlates strongly with the oracle upper bound performance. We also demonstrate large consistent performance gains for prompt settings over a range of NLP tasks.
  arXiv  Detail & Related papers  (2023-09-10T10:57:41Z)
• Generating Sequences by Learning to Self-Correct [64.0249217590888]
  Self-Correction decouples an imperfect base generator from a separate corrector that learns to iteratively correct imperfect generations. We show that Self-Correction improves upon the base generator in three diverse generation tasks.
  arXiv  Detail & Related papers  (2022-10-31T18:09:51Z)
• Measuring and Narrowing the Compositionality Gap in Language Models [116.5228850227024]
  We measure how often models can correctly answer all sub-problems but not generate the overall solution. We present a new method, self-ask, that further improves on chain of thought.
  arXiv  Detail & Related papers  (2022-10-07T06:50:23Z)
• Estimating Confidence of Predictions of Individual Classifiers and Their Ensembles for the Genre Classification Task [0.0]
  Genre identification is a subclass of non-topical text classification. Nerve models based on pre-trained transformers, such as BERT or XLM-RoBERTa, demonstrate SOTA results in many NLP tasks.
  arXiv  Detail & Related papers  (2022-06-15T09:59:05Z)
• Dealing with Disagreements: Looking Beyond the Majority Vote in Subjective Annotations [6.546195629698355]
  We investigate the efficacy of multi-annotator models for subjective tasks. We show that this approach yields same or better performance than aggregating labels in the data prior to training. Our approach also provides a way to estimate uncertainty in predictions, which we demonstrate better correlate with annotation disagreements than traditional methods.
  arXiv  Detail & Related papers  (2021-10-12T03:12:34Z)
• AvgOut: A Simple Output-Probability Measure to Eliminate Dull Responses [97.50616524350123]
  We build dialogue models that are dynamically aware of what utterances or tokens are dull without any feature-engineering. The first model, MinAvgOut, directly maximizes the diversity score through the output distributions of each batch. The second model, Label Fine-Tuning (LFT), prepends to the source sequence a label continuously scaled by the diversity score to control the diversity level. The third model, RL, adopts Reinforcement Learning and treats the diversity score as a reward signal.
  arXiv  Detail & Related papers  (2020-01-15T18:32:06Z)

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.