Datasets:

reasoning-degeneration-dev
/

gepa-abstract-reasoning-rlm

run_id string	iteration int64	cumulative_tokens int64	best_val_score float64	accepted bool	candidate_prompt string	best_prompt string
fixed_iter_rlm_k3	1	59,604	0.266667	false	You are a helpful assistant. You are given a question and you need to answer it. The answer should be given at the end of your response in exactly the format '### <final answer>'	You are a helpful assistant. You are given a question and you need to answer it. The answer should be given at the end of your response in exactly the format '### <final answer>'
fixed_iter_rlm_k3	2	118,346	0.266667	false	You are a helpful assistant. You are given a question and you need to answer it. The answer should be given at the end of your response in exactly the format '### <final answer>'	You are a helpful assistant. You are given a question and you need to answer it. The answer should be given at the end of your response in exactly the format '### <final answer>'
fixed_iter_rlm_k3	3	156,889	0.266667	false	You are a helpful assistant. You are given a question and you need to answer it. The answer should be given at the end of your response in exactly the format '### <final answer>'	You are a helpful assistant. You are given a question and you need to answer it. The answer should be given at the end of your response in exactly the format '### <final answer>'
fixed_iter_rlm_k3	4	203,508	0.266667	false	You are a helpful assistant. You are given a question and you need to answer it. The answer should be given at the end of your response in exactly the format '### <final answer>'	You are a helpful assistant. You are given a question and you need to answer it. The answer should be given at the end of your response in exactly the format '### <final answer>'
fixed_iter_rlm_k3	5	233,490	0.266667	false	You are a helpful assistant. You are given a question and you need to answer it. The answer should be given at the end of your response in exactly the format '### <final answer>'	You are a helpful assistant. You are given a question and you need to answer it. The answer should be given at the end of your response in exactly the format '### <final answer>'

gepa-abstract-reasoning-rlm

GEPA prompt optimization experiment on AIME math problems.

Task LM: openai/gpt-4.1-mini | Reflection LM: openai/gpt-5 | Reflection Mode: abstract_reasoning | Last updated: 2026-02-27 01:23 UTC

Results

Run	Method	k	Mode	Val Score	Test Acc	Tokens	Cost	Time
fixed_rlm_k3	rlm	3	abstract_reasoning	26.67%	N/A	233,490	$0.0000	2279s

Learning Curves

Experiment Config

{
  "script_name": "run_experiment.py",
  "model": "openai/gpt-4.1-mini",
  "reflection_lm": "openai/gpt-5",
  "reflection_mode": "abstract_reasoning",
  "hyperparameters": {
    "task_lm": "openai/gpt-4.1-mini",
    "reflection_lm": "openai/gpt-5",
    "reflection_mode": "abstract_reasoning",
    "seed_prompt": "You are a helpful assistant. You are given a question and you need to answer it. The answer should b..."
  },
  "input_datasets": [
    "AI-MO/aimo-validation-aime",
    "MathArena/aime_2025"
  ],
  "description": "GEPA prompt optimization experiment on AIME (reflection_mode=abstract_reasoning)",
  "num_runs": 1,
  "last_updated": "2026-02-27 01:23 UTC"
}

Dataset Configs

This repo contains multiple configs (subsets). Load them with:

from datasets import load_dataset

results    = load_dataset("reasoning-degeneration-dev/gepa-abstract-reasoning-rlm", "results", split="train")
traces     = load_dataset("reasoning-degeneration-dev/gepa-abstract-reasoning-rlm", "traces", split="train")
val_traces = load_dataset("reasoning-degeneration-dev/gepa-abstract-reasoning-rlm", "val_traces", split="train")
curves     = load_dataset("reasoning-degeneration-dev/gepa-abstract-reasoning-rlm", "curves", split="train")
rlm_call_traces = load_dataset("reasoning-degeneration-dev/gepa-abstract-reasoning-rlm", "rlm_call_traces", split="train")

Config	Description
`results`	Per-run summary: method, k, val/test scores, tokens, cost, optimized prompt
`traces`	Per-example test set traces with model responses and scores
`val_traces`	Per-iteration validation traces with the prompt used at each step
`curves`	Per-iteration learning curves: tokens, scores, candidate prompts
`rlm_call_traces`	Hierarchical LLM call traces: every RLM iteration, sub-call, prompt, and response

State file: state.json

For experiment design details, see EXPERIMENTS-DESIGN.md

Downloads last month: 59

Number of rows:

Total file size:

1.58 MB