mlsanityData to 14 Oct 2026

Direct-Inverse Prompting: Analyzing LLMs' Discriminative Capacity in Self-Improving Generation

arXiv 2407.11017 ↗First posted 27 Jun 2024cs.CL cs.AI cs.LGcited by 0

On the map: Synthetic data and self-improvement for LLMs → Self-rewarding alignment and self-training

Abstract (arXiv)

Mainstream LLM research has primarily focused on enhancing their generative capabilities. However, even the most advanced LLMs experience uncertainty in their outputs, often producing varied results on different runs or when faced with minor changes in input, despite no substantial change in content. Given multiple responses from the same LLM to the same input, we advocate leveraging the LLMs' discriminative capability to reduce this generative uncertainty, aiding in identifying the correct answers. Specifically, we propose and analyze three discriminative prompts: direct, inverse, and hybrid, to explore the potential of both closed-source and open-source LLMs in self-improving their generative performance on two benchmark datasets. Our insights reveal which discriminative prompt is most promising and when to use it. To our knowledge, this is the first work to systematically analyze LLMs' discriminative capacity to address generative uncertainty.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Self-rewarding alignment and self-training, one of 13 directions in Synthetic data and self-improvement for LLMs. Builds on 12 of its 17 references that are in the map. First posted on arXiv 27 Jun 2024.

Builds on

12 of 17 references.

References · 12

  1. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,708
  2. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,371
  3. 2023Teaching Large Language Models to Self-Debugcited by 988
  4. 2023CRITIC: Large Language Models Can Self-Correct with Tool-Interactive Critiquingcited by 783
  5. 2019MathQA: Towards Interpretable Math Word Problem Solving with Operation-Based Formalismscited by 662
  6. 2022RARR: Researching and Revising What Language Models Say, Using Language Modelscited by 287
  7. 2023REFINER: Reasoning Feedback on Intermediate Representationscited by 205
  8. 2023Universal Self-Consistency for Large Language Model Generationcited by 203
  9. 2024Escape Sky-high Cost: Early-stopping Self-Consistency for Multi-step Reasoningcited by 108
  10. 2023Shepherd: A Critic for Language Model Generationcited by 97
  11. 2023Benchmarking Generation and Evaluation Capabilities of Large Language Models for Instruction Controllable Summarizationcited by 69
  12. 2023Learning and Leveraging Verifiers to Improve Planning Capabilities of Pre-trained Language Modelscited by 10

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024SELF-[IN]CORRECT: LLMs Struggle with Discriminating Self-Generated Responses
  2. 2025LLMs Can Generate a Better Answer by Aggregating Their Own Responses
  3. 2024Self-Instructed Derived Prompt Generation Meets In-Context Learning: Unlocking New Potential of Black-Box LLMs
  4. 2026Evaluating LLMs as Human Surrogates in Controlled Experiments
  5. 2024Toward Self-Improvement of LLMs via Imagination, Searching, and Criticizing
  6. 2025Flaw or Artifact? Rethinking Prompt Sensitivity in Evaluating LLMs
  7. 2024Understanding the Dark Side of LLMs' Intrinsic Self-Correction
  8. 2026Not the Example, but the Process: How Self-Generated Examples Enhance LLM Reasoning