Pith. sign in

REVIEW 7 cited by

TEMPERA: Test-Time Prompting via Reinforcement Learning

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2211.11890 v1 pith:BOC52NC6 submitted 2022-11-21 cs.CL cs.AI

classification cs.CLcs.AI
keywords promptdesignlearningmethodstemperaachievescomparedediting
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Careful prompt design is critical to the use of large language models in zero-shot or few-shot learning. As a consequence, there is a growing interest in automated methods to design optimal prompts. In this work, we propose Test-time Prompt Editing using Reinforcement learning (TEMPERA). In contrast to prior prompt generation methods, TEMPERA can efficiently leverage prior knowledge, is adaptive to different queries and provides an interpretable prompt for every query. To achieve this, we design a novel action space that allows flexible editing of the initial prompts covering a wide set of commonly-used components like instructions, few-shot exemplars, and verbalizers. The proposed method achieves significant gains compared with recent SoTA approaches like prompt tuning, AutoPrompt, and RLPrompt, across a variety of tasks including sentiment analysis, topic classification, natural language inference, and reading comprehension. Our method achieves 5.33x on average improvement in sample efficiency when compared to the traditional fine-tuning methods.

Discussion (0). Sign in to comment.

Forward citations

Cited by 7 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. TAPR: Enhancing LLM Performance with a Task-Aware Prompt Rewriter

    cs.AI 2026-07 conditional novelty 6.0 of 10

    A small LLM trained with GRPO and LLM-judge rewards rewrites simple prompts into more effective ones, improving question-answering and arithmetic accuracy over base prompts while giving mixed, often negligible gains o...

  2. TRPrompt: Bootstrapping Query-Aware Prompt Optimization from Textual Rewards

    cs.CL 2025-07 conditional novelty 6.0 of 10

    TRPrompt trains an 8B prompt model directly on natural-language textual rewards and reports the highest accuracies on GSMHard and MATH among the compared methods.

  3. DEMONSTRATE: Zero-shot Language to Robotic Control via Multi-task Demonstration Learning

    cs.RO 2025-07 conditional novelty 6.0 of 10

    DEMONSTRATE learns a zero-shot mapping from natural-language embeddings to MPC cost parameters from demonstrations, achieving tabletop manipulation success rates comparable to prior LLM-based pipelines.

  4. Grammar-Guided Evolutionary Search for Discrete Prompt Optimisation

    cs.CL 2025-07 conditional novelty 6.0 of 10

    A grammar-guided evolutionary search that composes prompt edits outperformed PromptWizard, OPRO, and RL-Prompt on small LLMs across four domain-specific tasks.

  5. Stabilizing Black-Box Prompt Optimization with Textual Regularization and Signal Aggregation

    cs.LG 2025-07 conditional novelty 6.0 of 10

    TRAS adds success-based textual regularization and Monte Carlo signal aggregation to black-box prompt optimization, improving accuracy and reducing instruction loss when moving prompts across models.

  6. In-Context Learning as Implicit Policy Gradient

    cs.LG 2026-07 conditional novelty 5.0 of 10

    Score-conditioned in-context learning is structurally analogous to a REINFORCE policy-gradient update in a simplified hidden-state model, with attention performing score-weighted aggregation and a bounded KL shift.

  7. Evolutionary Computation and Large Language Models: A Survey of Methods, Synergies, and Applications

    cs.NE 2025-05 conditional novelty 4.0 of 10

    A survey that maps bidirectional synergies between evolutionary computation and large language models and proposes a taxonomy plus research gaps.

Pith tools