List your company

Table Reasoning Accuracy on TableBench

63.48AccuracySERO
27.901637.138346.37555.6117May 27, 2026May 28, 2026

Evaluation results

MethodLinks
SEROBackbone=Gemini-2.5-flash-lite2026.0563.48—
Static Role OrchestrationBackbone=Gemini-2.5-flash-lite2026.0561.6—
Static DAG MASBackbone=Gemini-2.5-flash-lite2026.0560.68—
Random Role EvolutionBackbone=Gemini-2.5-flash-lite2026.0560.64—
CoTBackbone=Gemini-2.5-flash-lite2026.0559—
SC@3Backbone=Gemini-2.5-flash-lite2026.0558.5—
WorkflowBackbone=Gemini-2.5-flash-lite2026.0557.91—
14B OnlyBackbone=Qwen3-Instruct, Extra Tokens=–2026.0554.8542.4
SEROBackbone=GPT-4o-mini2026.0554.06—
14B OnlyBackbone=DeepSeek-R1-Distill-Qwen, Extra Tokens=–2026.0553.6540.5
Static Role OrchestrationBackbone=GPT-4o-mini2026.0552.43—
Static DAG MASBackbone=GPT-4o-mini2026.0552.3—
EcoTabBackbone=Qwen3-Instruct, Extra Tokens=✗2026.0551.7728.1
Random Role EvolutionBackbone=GPT-4o-mini2026.0551.76—
GlimpRouterBackbone=Qwen3-Instruct, Extra Tokens=✓2026.0550.5230.1
EcoTabBackbone=DeepSeek-R1-Distill-Qwen, Extra Tokens=✗2026.0550.4127.4
STEERBackbone=Qwen3-Instruct, Extra Tokens=✗2026.0550.3329.4
SpecReasonBackbone=Qwen3-Instruct, Extra Tokens=✓2026.0550.0229.9
SpecCoTBackbone=Qwen3-Instruct, Extra Tokens=✓2026.0549.8729.7
GlimpRouterBackbone=DeepSeek-R1-Distill-Qwen, Extra Tokens=✓2026.0549.5428.1
SEROBackbone=Qwen3-8b2026.0549.37—
STEERBackbone=DeepSeek-R1-Distill-Qwen, Extra Tokens=✗2026.0549.2229.1
RSDBackbone=Qwen3-Instruct, Extra Tokens=✓2026.0549.2130.4
WorkflowBackbone=GPT-4o-mini2026.0549.08—
SpecReasonBackbone=DeepSeek-R1-Distill-Qwen, Extra Tokens=✓2026.0548.8928.4
RSDBackbone=DeepSeek-R1-Distill-Qwen, Extra Tokens=✓2026.0548.8128.6
SpecCoTBackbone=DeepSeek-R1-Distill-Qwen, Extra Tokens=✓2026.0548.7428.3
Random Role EvolutionBackbone=Qwen3-8b2026.0548.32—
SC@3Backbone=GPT-4o-mini2026.0547.99—
Static Role OrchestrationBackbone=Qwen3-8b2026.0547.95—
CoTBackbone=GPT-4o-mini2026.0547.24—
RandomBackbone=Qwen3-Instruct, Extra Tokens=✗2026.0546.1236.7
WorkflowBackbone=Qwen3-8b2026.0544.43—
Static DAG MASBackbone=Qwen3-8b2026.0544.05—
RandomBackbone=DeepSeek-R1-Distill-Qwen, Extra Tokens=✗2026.0544.0336.4
1.7B OnlyBackbone=Qwen3-Instruct, Extra Tokens=–2026.0542.556.12
1.7B OnlyBackbone=DeepSeek-R1-Distill-Qwen, Extra Tokens=–2026.0542.556.12
CoTBackbone=Qwen3-8b2026.0529.44—
SC@3Backbone=Qwen3-8b2026.0529.27—