Fact Chaining & Relational Reasoning on Long-context benchmarks (8k-128k context) benchmark leaderboard · SOTA2 Research

Fact Chaining & Relational Reasoning on Long-context benchmarks (8k-128k context)

52.8Accuracy (8k Context)DCA

Evaluation Results

Method	Links
DCA2026.03		52.8	47	46	40.8	30
Base2026.03		52.8	47.2	44	27.6	0.4
SelfExtend2026.03		52.2	47.6	41.8	38.2	34.6
DCA2026.03		51.6	47.8	44.8	39.8	35.2
Base2026.03		51.6	47.8	43	41.8	21.4
FlexPrefill2026.03		47.6	42.6	37.8	33.4	35.8
CLP2026.03		47.4	43.6	40	31.4	38.6
L2A2026.03		47.2	47.6	47.6	38.6	37.8
S^2-attn2026.03		45.2	43.8	35.2	38.2	39.2
FlexPrefill2026.03		44.8	39	39	34.2	31
L2A (Ours)2026.03		44.6	42.2	40.6	33.4	40.4
CLP2026.03		43.6	39.8	35.8	28.2	29.4
S^2-attn2026.03		43.4	40.2	36.2	31.6	26.4
CLP2026.03		43.4	43.6	46.6	37.8	35.6
FlexPrefill2026.03		43.2	37.6	36	32	27.4
NSA2026.03		42.2	39.2	31.2	24	17.6
L2A2026.03		41.4	36	—	29.2	30
NSA2026.03		40.4	30.6	23	18	11
Base2026.03		38.2	35.2	31	24	13
DCA2026.03		37	35	35.6	27.8	23.4
SelfExtend2026.03		37	35.6	34	26.2	23.4