Share your thoughts, 1 month free Claude Pro on usSee more

Time-series Reasoning on TSRBench

64.3Percept Score

TIMECLAW

Updated 1mo ago

Evaluation Results

Method	Links
TIMECLAW 2026.06		64.3	43.9	59.3	33.3	49.8
Qwen3-235B-A22B 2026.06		63.8	40.9	35.6	32.5	42.1
MiniCPM-V-4.5-8B 2026.06		61.3	26.5	39	27.9	35.9
InternVL3.5-38B 2026.06		60.3	38.1	32.3	32.3	39.5
InternVL3.5-8B 2026.06		59.3	35.6	39.4	28.2	39.5
MiMo-VL-7B-RL 2026.06		59.1	33.1	33.2	30.9	37.2
Gemma3-27B-it 2026.06		57	33.7	38.3	31.8	38.6
Qwen2.5-72B 2026.06		55.6	39.7	43.9	32.2	42.4
TimeOmni-1-7B 2026.06		54.7	31	36.5	32.2	36.6
Gemma3-12B-it 2026.06		54.3	30.6	37.7	32.9	36.8
Qwen3-8B 2026.06		52.6	29.1	47.4	32	38.3
TS-Reasoner-7B 2026.06		52.1	31.3	39.5	27.6	36.4
ChatTS-14B 2026.06		50.7	29.3	30.2	28.5	33
Qwen2.5-7B 2026.06		49.6	28.3	34.8	28.5	33.7
Phi4-Multimodal-8B 2026.06		49	26	30	32.1	31.9
Qwen3-1.7B 2026.06		47	27.7	48.7	28.2	36.6
InternVL3.5-1B 2026.06		45.9	27.8	38.6	28.2	33.8
Qwen2.5-3B 2026.06		45.4	27.3	40.4	23.3	33.2
OpenTSLM-tsqa-sp-3B 2026.06		39.7	26.8	33.5	28.5	31