Pith. sign in

REVIEW 8 cited by

Planning In Natural Language Improves LLM Search For Code Generation

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 2409.03733 v2 pith:ASBVJQKM submitted 2024-09-05 cs.LG cs.AIcs.CL

Planning In Natural Language Improves LLM Search For Code Generation

classification cs.LG cs.AIcs.CL
keywords searchplansearchlanguagecodediversemodelsnaturalpass
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

While scaling training compute has led to remarkable improvements in large language models (LLMs), scaling inference compute has not yet yielded analogous gains. We hypothesize that a core missing component is a lack of diverse LLM outputs, leading to inefficient search due to models repeatedly sampling highly similar, yet incorrect generations. We empirically demonstrate that this lack of diversity can be mitigated by searching over candidate plans for solving a problem in natural language. Based on this insight, we propose PlanSearch, a novel search algorithm which shows strong results across HumanEval+, MBPP+, and LiveCodeBench (a contamination-free benchmark for competitive coding). PlanSearch generates a diverse set of observations about the problem and then uses these observations to construct plans for solving the problem. By searching over plans in natural language rather than directly over code solutions, PlanSearch explores a significantly more diverse range of potential solutions compared to baseline search methods. Using PlanSearch on top of Claude 3.5 Sonnet achieves a state-of-the-art pass@200 of 77.0% on LiveCodeBench, outperforming both the best score achieved without search (pass@1 = 41.4%) and using standard repeated sampling (pass@200 = 60.6%). Finally, we show that, across all models, search algorithms, and benchmarks analyzed, we can accurately predict performance gains due to search as a direct function of the diversity over generated ideas. Code can be found at https://github.com/scaleapi/plansearch.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 8 Pith papers

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

  1. Falsification, Not Exposure: An Internally Preregistered Placebo-Controlled Decomposition of Self-Repair Feedback in Frozen Small Code Models

    cs.SE 2026-06 unverdicted novelty 7.0

    Preregistered placebo-controlled decomposition shows external executable counterexamples drive self-repair gains in small code models more than re-exposure or self-critique.

  2. Syntax Without Semantics: Teaching Large Language Models to Code in an Unseen Language

    cs.CL 2026-05 unverdicted novelty 7.0

    Fine-tuning LLMs on an unseen language teaches syntax but fails to transfer semantic competence, leaving Python with up to a 19% performance advantage and no tested intervention closing the gap.

  3. AdverMCTS: Combating Pseudo-Correctness in Code Generation via Adversarial Monte Carlo Tree Search

    cs.SE 2026-04 unverdicted novelty 7.0

    AdverMCTS frames code generation as a minimax game where an attacker evolves tests to expose flaws in solver-generated code, yielding more robust outputs than static-test baselines.

  4. OPD-Evolver: Cultivating Holistic Agent Evolver via On-Policy Distillation

    cs.CL 2026-06 unverdicted novelty 6.0

    OPD-Evolver uses on-policy self-distillation in fast interaction and slow attribution loops to build agents with holistic memory competence, outperforming prior systems by up to 11.5% and allowing a 9B model to compet...

  5. Syntax Without Semantics: Teaching Large Language Models to Code in an Unseen Language

    cs.CL 2026-05 conditional novelty 6.0

    Fine-tuning LLMs on a novel, syntax-friendly but feature-poor language teaches parsing without closing the gap between PyLang and Python pass rates.

  6. Evaluation-driven Scaling for Scientific Discovery

    cs.LG 2026-04 unverdicted novelty 6.0

    SimpleTES scales test-time evaluation in LLMs to discover state-of-the-art solutions on 21 scientific problems across six domains, outperforming frontier models and optimization pipelines with examples like 2x faster ...

  7. Code as Agent Harness

    cs.CL 2026-05 accept novelty 5.0

    A survey that organizes existing work on LLM-based agents around code as the central harness, structured in three layers of interfaces, mechanisms, and multi-agent scaling, with applications across domains and listed ...

  8. Muse Spark Safety & Preparedness Report

    cs.CY 2026-05 unverdicted novelty 2.0

    Meta's safety report states that Muse Spark meets acceptable risk thresholds for release after mitigations reduced elevated pre-mitigation risks in chemical and biological domains.