HyperAI超神経

Table Based Fact Verification On Tabfact

評価指標

Test
Val

評価結果

このベンチマークにおける各モデルのパフォーマンス結果

モデル名
Test
Val
Paper TitleRepository
TAPAS-Large classifier with Counterfactual + Synthetic pre-training81.081.0Understanding tables with intermediate pre-training
PASTA89.389.2PASTA: Table-Operations Aware Fact Verification via Sentence-Table Cloze Pre-training-
T5-3b(UnifiedSKG)83.6883.97UnifiedSKG: Unifying and Multi-Tasking Structured Knowledge Grounding with Text-to-Text Language Models
Table-BERT-Horizontal-T+F-Template65.1266.1TabFact: A Large-scale Dataset for Table-based Fact Verification
ARTEMIS-DA93.1-ARTEMIS-DA: An Advanced Reasoning and Transformation Engine for Multi-Step Insight Synthesis in Data Analytics-
Binder86.0-Binding Language Models in Symbolic Languages
TabSQLify (col+row)79.5-TabSQLify: Enhancing Reasoning Capabilities of LLMs Through Table Decomposition-
ReasTAP-Large84.984.6ReasTAP: Injecting Table Reasoning Skills During Pre-training via Synthetic Reasoning Examples-
Chain-of-Table86.61-Chain-of-Table: Evolving Tables in the Reasoning Chain for Table Understanding
Tab-PoT85.77-Efficient Prompting for LLM-based Generative Internet of Things-
TAPEX-Large84.284.6TAPEX: Table Pre-training via Learning a Neural SQL Executor
NormTab (Targeted) + SQL68.90-NormTab: Improving Symbolic Reasoning in LLMs Through Tabular Data Normalization
BERT classifier w/o Table50.550.9TabFact: A Large-scale Dataset for Table-based Fact Verification
Dater93.0-Large Language Models are Versatile Decomposers: Decompose Evidence and Questions for Table-based Reasoning
Salience-aware TAPAS82.182.7Table-based Fact Verification with Salience-aware Learning
0 of 15 row(s) selected.