Multi-step Reasoning

Benchmarks

Dataset Name	SOTA Method	Metric
ScienceQA	SceneAlign	Accuracy92.72	35	4mo ago
StrategyQA	Qwen3-8B + SFT + WeMask(TF)	Accuracy66.99	20	17d ago
TriviaQA	LAR	Task Performance80.09	14	2mo ago
SVAMP	eMoT	Accuracy94	11	25d ago
StrategyQA (test)	Qwen3-4B + SFT + WeMask(TF)	Accuracy64.63	11	2mo ago
Bamboogle auto-eval (test)	Self-improvement, 2nd gen	Mean Accuracy76.1	10	4mo ago
GSM8K (test)	SLR	Pass@132.2	9	4mo ago
Bamboogle (test)	CAHL	Accuracy75.2	9	1mo ago
Big-Bench Hard	Ouro-1.4B-R4	Accuracy71.02	8	1mo ago
CLEVR-Puzzle (test)	NeSyCoCo	Accuracy95	7	4mo ago
GSM Hard	eMoT	Accuracy71.5	5	1mo ago
GAIA (test)	Skill-R1 (GRPO, Qwen3-4b)	Level 1 Accuracy31	4	2mo ago
MuSR	n=64 (RL)	Murder Mystery Score57.36	4	2mo ago
Musique 16k ~ 20k	Absorber LLM	Macro Accuracy29.5	4	3mo ago
Musique 12k ~ 16k	Absorber LLM	Accuracy (macro)31.6	4	3mo ago
Musique 8k ~ 12k	TTT	Macro Accuracy34.3	4	3mo ago
BamTwoogle (test)	ReST meets ReAct	Accuracy74	4	4mo ago
Checkmate	eMoT	Accuracy90	3	1mo ago
WordSort	BoT	Accuracy100	3	1mo ago
Game24	eMoT	Accuracy100	3	1mo ago
MUSR Teams	Static Workflow	Accuracy61	3	1mo ago
MUSR Objects	Static Workflow	Accuracy58	3	1mo ago
MUSR Murder	Static Workflow	Accuracy68	3	1mo ago
multi-step reasoning tasks	EmbodiedAct	Average Score61.6	3	4mo ago
BabyAI-Text UnlockLocal (New Object)	AEC	Success Rate95	2	26d ago

Showing 25 of 26 rows