Search-SWE
GPU TASK Valid result

ChunQiu Temporal Retrieval

GPT-6 Luna · Codex (v0.164.0) · reasoning effort max

Task score (%)1.7
Input Tokens13,553,651including 13,278,544 cached
Output Tokens184,124
Agent time49min
Cost$0.28
Recorded steps86

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
2026-10-24T07:21:48.815174Z
Coding finished
2026-10-24T08:11:07.424679Z
Input (includes cache)
13,553,651
Cached input (subset)
13,278,544
Cache writes (subset)
274,340
Output
184,124
Reasoning (reported separately; not added)
135,098
Recorded responses
83

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": null,
  "coding_commit": null,
  "effort": "max",
  "model": "openai/gpt-6-luna",
  "provider": "OpenRouter",
  "resources": null,
  "run_mode": "saved_outputs_production_rescore",
  "task_commit": "1d73c6f9c4a90ba922b7572d9395a6f40661bbb2",
  "verifier_commit": "1d73c6f9c4a90ba922b7572d9395a6f40661bbb2",
  "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
chunqiu-temporal-retrieval-gpt-6-luna-max-d1be8212ebca
Source trajectory SHA-256
4c6b521c1a3ef7b1bf294ef7ba69f5a0b061a255a1e04bd3e7ddf2255d82d32d
Final reward SHA-256
178436c258606735f96b0262919e7f0d7898639c9d8feeb980ad3450e3f82e92