Skip to main content
AI/MLparcadei

llm-tuning-patterns

LLM Tuning Patterns

Stars
3,795
Source
parcadei/Continuous-Claude-v3
Updated
2026-01-26
Slug
parcadei--Continuous-Claude-v3--llm-tuning-patterns
View on GitHubRaw SKILL.md

// install — copy + paste into any project

mkdir -p .claude/skills && curl -fsSL https://raw.githubusercontent.com/parcadei/Continuous-Claude-v3/HEAD/.claude/skills/llm-tuning-patterns/SKILL.md -o .claude/skills/llm-tuning-patterns.md

Drops the SKILL.md into .claude/skills/llm-tuning-patterns.md. Works with Claude Code, Cursor, and any agent that loads SKILL.md files from .claude/skills/.

LLM Tuning Patterns

Evidence-based patterns for configuring LLM parameters, based on APOLLO and Godel-Prover research.

Pattern

Different tasks require different LLM configurations. Use these evidence-based settings.

Theorem Proving / Formal Reasoning

Based on APOLLO parity analysis:

Parameter Value Rationale
max_tokens 4096 Proofs need space for chain-of-thought
temperature 0.6 Higher creativity for tactic exploration
top_p 0.95 Allow diverse proof paths

Proof Plan Prompt

Always request a proof plan before tactics:

Given the theorem to prove:
[theorem statement]

First, write a high-level proof plan explaining your approach.
Then, suggest Lean 4 tactics to implement each step.

The proof plan (chain-of-thought) significantly improves tactic quality.

Parallel Sampling

For hard proofs, use parallel sampling:

  • Generate N=8-32 candidate proof attempts
  • Use best-of-N selection
  • Each sample at temperature 0.6-0.8

Code Generation

Parameter Value Rationale
max_tokens 2048 Sufficient for most functions
temperature 0.2-0.4 Prefer deterministic output

Creative / Exploration Tasks

Parameter Value Rationale
max_tokens 4096 Space for exploration
temperature 0.8-1.0 Maximum creativity

Anti-Patterns

  • Too low tokens for proofs: 512 tokens truncates chain-of-thought
  • Too low temperature for proofs: 0.2 misses creative tactic paths
  • No proof plan: Jumping to tactics without planning reduces success rate

Source Sessions

  • This session: APOLLO parity - increased max_tokens 512->4096, temp 0.2->0.6
  • This session: Added proof plan prompt for chain-of-thought before tactics