datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
gepa-rlm-exp-algorithmic-20260219-191545
gepa-rlm-exp-algorithmic-20260219-191545
GEPA prompt optimization experiment on AIME math problems.
Task LM: openai/gpt-4.1-mini | Reflection LM: openai/gpt-5 | Reflection Mode: algorithmic | Last updated: 2026-02-19 21:34 UTC
Results
Run
Method
k
Mode
Val Score
Test Acc
Tokens
Cost
Time
fixed_rlm_k20
rlm
20
algorithmic
48.89%
26.67%
932,391
$0.0000
5209s
Learning Curves
Experiment Config
{
"script_name": "run_experiment.py"… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/gepa-rlm-exp-algorithmic-20260219-191545.gepa-exp-algorithmic-vanilla-20260220-083526
gepa-exp-algorithmic-vanilla-20260220-083526
GEPA prompt optimization experiment on AIME math problems.
Task LM: openai/gpt-4.1-mini | Reflection LM: openai/gpt-5 | Reflection Mode: algorithmic | Last updated: 2026-02-20 21:20 UTC
Results
Run
Method
k
Mode
Val Score
Test Acc
Tokens
Cost
Time
fixed_vanilla_k3
vanilla
3
algorithmic
35.56%
46.67%
77,464
$0.2182
3126s
Learning Curves
Experiment Config
{
"script_name":… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/gepa-exp-algorithmic-vanilla-20260220-083526.algorithmic-sft-training-data-v1
algorithmic-sft-training-data-v1
Algorithmic SFT training data: deterministic step-by-step traces for 5 domains (countdown, formal_logic, long_arithmetic, conlang_morphology, cellular_automata) across multiple algorithm variants. Programmatically generated — no LLM involved.
Dataset Info
Rows: 63000
Columns: 8
Columns
Column
Type
Description
question
Value('string')
The problem statement presented to the model
answer
Value('string')
The correct… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/algorithmic-sft-training-data-v1.gepa-exp-algorithmic-rlm-20260220-083526
gepa-exp-algorithmic-rlm-20260220-083526
GEPA prompt optimization experiment on AIME math problems.
Task LM: openai/gpt-4.1-mini | Reflection LM: openai/gpt-5 | Reflection Mode: algorithmic | Last updated: 2026-02-20 21:25 UTC
Results
Run
Method
k
Mode
Val Score
Test Acc
Tokens
Cost
Time
fixed_rlm_k3
rlm
3
algorithmic
37.78%
44.00%
400,201
$0.0000
3483s
Learning Curves
Experiment Config
{
"script_name": "run_experiment.py"… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/gepa-exp-algorithmic-rlm-20260220-083526.algorithmic-sft-training-configs-v1
algorithmic-sft-training-configs-v1
LlamaFactory training configs. All cutoff_len=32768. Countdown configs use new equation-answer format.
Dataset Info
Rows: 17
Columns: 7
Columns
Column
Type
Description
config_name
Value('string')
YAML filename
domain
Value('string')
No description provided
is_distillation
Value('bool')
No description provided
yaml_content
Value('string')
Full YAML config
model_name
Value('string')
No description provided… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/algorithmic-sft-training-configs-v1.algorithmic-sft-sharegpt-training-v1
algorithmic-sft-sharegpt-training-v1
Exact LlamaFactory training data. Countdown uses equation-answer format with Step 1:.... Other domains use Answer: X. All wrapped in tags.
Dataset Info
Rows: 82903
Columns: 4
Columns
Column
Type
Description
conversations
List({'from': Value('string'), 'value': Value('string')})
ShareGPT — literal LlamaFactory input
source_file
Value('string')
JSON filename → dataset_info.json
model_type
Value('string')… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/algorithmic-sft-sharegpt-training-v1.algorithmic-sft-eval-sets-v1
algorithmic-sft-eval-sets-v1
Evaluation sets for algorithmic SFT experiment: test (1000/domain), val (200/domain), harder variant (500/domain), structural OOD (500/domain). 5 domains total.
Dataset Info
Rows: 11000
Columns: 9
Columns
Column
Type
Description
question
Value('string')
The problem statement
answer
Value('string')
The correct answer
sft_trace
Value('null')
Reference algorithmic trace (golden solution)
difficulty
Value('int64')… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/algorithmic-sft-eval-sets-v1.algorithmic-sft-distillation-training-data-v1
algorithmic-sft-distillation-training-data-v1
QwQ-32B distillation training data for 5 algorithmic domains. Correct responses filtered by collect_distill_results.py with truncation rejection. KNOWN ISSUE: countdown domain has 74.2% QwQ repetition loops (model repeats \boxed{} answer until hitting token limit). Countdown is being regenerated with v3 pipeline (32k tokens). Other 4 domains are clean (>99% quality).
Dataset Info
Rows: 24133
Columns: 6
Columns… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/algorithmic-sft-distillation-training-data-v1.reasoning-gym-dataset-algorithmic-tasksreasoninggym-algorithmicalgorithmic-sft-full-eval-v3
algorithmic-sft-full-eval-v3
Aggregate eval results: 16 models × 3 splits (test/harder/OOD). v3 re-evaluation at MAX_TOKENS=32768. Bootstrap 95% CI.
Dataset Info
Rows: 50
Columns: 14
Columns
Column
Type
Description
eval_name
Value('string')
Filename-derived eval identifier
domain
Value('string')
Task domain (countdown, formal_logic, long_arithmetic, cellular_automata, conlang_morphology)
variant
Value('string')
Model variant (algorithm name or… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/algorithmic-sft-full-eval-v3.