Benchmark Almanac

Karmendra Pandey

We track 3 benchmarks that Karmendra Pandey made or helped make.

0 in use 0 with public code 2 with a leaderboard here

Benchmarks

Name Kind Released Best score Status
Argus-Bench v2 Models screen inputs to a language model gateway. The benchmark checks detection rates, false positives, and agreement with the policy on severity. Jailbreaks and prompt injection Nov 2026 85.7% Not in cards
Cost Per Correct v2 Models answer tasks at three difficulty levels and repeat some with paraphrased instructions. Rules check correctness so the author can compare cost per correct answer. General assistants Nov 2026 $91 Not in cards
Cost Per Correct Models complete five tasks with verifiable answers. The evaluation measures the price of obtaining a correct answer. General assistants Oct 2026 Not in cards