CPU TASK Valid result
ICSI Meeting Memory
GPT-6 Luna · Codex · reasoning effort max
Task score (%)0.0
Input Tokens43,594,608including 43,104,028 cached
Output Tokens306,495
Agent time1h 37min
Cost$1.06
Recorded steps150
Cost is estimated from recorded agent tokens at official API rates in USD (2026-11-01; DeepSeek off-peak pricing).
Loading trajectory…
Select this tab to browse the saved files.
Coding time & token usage
- Original agent completion
- normal
- Coding started
- 2026-10-24T03:26:23.876051Z
- Coding finished
- 2026-10-24T05:03:28.139684Z
- Input (includes cache)
- 43,594,608
- Cached input (subset)
- 43,104,028
- Cache writes (subset)
- 490,139
- Output
- 306,495
- Reasoning (reported separately; not added)
- 221,143
- Recorded responses
- 147
Input includes cached input; output is counted once. Auxiliary reasoning counts are shown as recorded. Regrading does not repeat coding usage.
Select this tab to browse the saved files.
Configuration
Evaluation settings summary
{
"agent": "Codex",
"agent_timeout_sec": 7200,
"coding_commit": null,
"effort": "max",
"model": "openai/gpt-6-luna",
"provider": "OpenRouter",
"resources": null,
"run_mode": "official_harbor_regrade",
"task_commit": null,
"verifier_commit": "1537d01c332d4c69160e7106757ac32602ba79b3",
"version": "0.164.0"
}Select this tab to browse the saved files.
Resolved configuration
Select this tab to browse the saved files.
Trial log
Select this tab to browse the saved files.
Exceptions
Select this tab to browse the saved files.
Select this tab to browse the saved files.
Trajectory downloads & provenance
Public text redacts private values and normalizes host paths. Source hashes identify the original records.
- Trial ID
icsi-meeting-memory-gpt-6-luna-max-9812ef66dc42- Source trajectory SHA-256
ad1e2720f322d80fd318743618e4094fa78db712e12fcb7dfbe72c49d50dfc49- Final reward SHA-256
b3cd28a46c3babda6f1364c913161a99b9b9d567bc269c8190960b7006ca80cd