Skip to content
NewA calmer, faster ZeroHour

TxBench benchmark for malicious transaction detection on EVM chains

Original title: TxBench: A Benchmark of Attack Transactions and Evidence-Backed Negatives on Seven EVM Chains

arXiv cs.CRby Tamer Abdelaziz ()1 min read
AI summary · x-ai/grok-4.7Summaries by Vani AI

TxBench includes 1012 attacks worth $868 million from 602 incidents

TxBench is a benchmark of attack transactions and evidence-backed negatives from seven EVM chains. It accepts 1012 attacks from 602 incidents worth USD 868.1 million. The five-step pipeline rebuilds executions from free sources. Zero-shot LLMs find at most 0.43 of attacks.

Key facts
  • 846 contract exploits and 166 key compromises
  • 98% of sampled attacks confirmed in blind review
  • Negatives sampled from chain traffic near attacks
  • Based on BlockSec incident library and EVM Hack Registry
Analyze this story with Vani AIImpact, who is affected and what defenders should do, in a new chat

From the original article

1 min read at arxiv.org

Researchers evaluate detectors of malicious smart-contract transactions on datasets that each paper assembles anew. Their positives come from incident trackers, whose citations mix attacks with setup steps and transactions that never ran, or from a detector's own output, which makes the evaluation circular; their negatives are ordinary transactions.

arXiv cs.CR

By Tamer Abdelaziz · arxiv.org

Read the full article at arXiv cs.CR

A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.