VeriLoop E2 evaluation evidence · inference-research

Fields

id

sources/veriloop-e2-evaluation-evidence

URL

https://huggingface.co/datasets/tsinghua-sigs-robot-lab/VeriLoop-E2-Evaluation-Evidence

Publisher

Tsinghua SIGS Robot Lab

Kind

benchmark

Published

2026-10-10

Accessed

2026-10-18

notes

Dataset of evaluation traces the model card cites as the source of its benchmark table. Subdirectories swe-bench-verified, swe-bench-pro, terminal-bench-2.1, gpqa-diamond, aime-2026 and livecodebench-v6. The .eval_results YAML on the model repo points each score at one of these subdirectories and carries no verified flag and no harness, seed or sample-count description, so the numbers stay vendor claims.

Notes

Dataset of evaluation traces the model card cites as the source of its benchmark table. Subdirectories swe-bench-verified, swe-bench-pro, terminal-bench-2.1, gpqa-diamond, aime-2026 and livecodebench-v6. The .eval_results YAML on the model repo points each score at one of these subdirectories and carries no verified flag and no harness, seed or sample-count description, so the numbers stay vendor claims.

Cross-references

This record references no other records.