PageSourceSearch

https://linzwcs.github.io/EvoPolicyGym/assets/js/7386cced.d5e5f867.js

js linzwcs.github.io collected 2026-10-03 09:32:54 UTC 3,979 bytes, 1 lines download raw bytes

1"use strict";(self.webpackChunkevopolicygym_site=self.webpackChunkevopolicygym_site||[]).push([[6338],{8339(e){e.exports=JSON.parse('{"manifest":{"schema":"evopolicygym/leaderboard-suite/v2","id":"ale-v1","slug":"ale-v1","status":"active","default":false,"version":"v1","experiment_version":"v0.3.0","title":{"en":"ALE Distribution","zh":"ALE Distribution"},"label":{"en":"ALE","zh":"ALE"},"description":{"en":"Independent Environment leaderboards for published Arcade Learning Environment experiments.","zh":"\u5df2\u516c\u5f00 Arcade Learning Environment \u5b9e\u9a8c\u7684\u72ec\u7acb Environment \u6392\u884c\u699c\u3002"},"profile":{"id":"ale-128-v1","label":{"en":"ALE 128-Episode profile","zh":"ALE 128-Episode \u914d\u7f6e"},"budget":128,"budget_unit":"training episodes / run"},"content":{"suite":{"en":"index.mdx","zh":"index.zh-CN.mdx"},"environment":{"en":"environment.mdx","zh":"environment.zh-CN.mdx"}},"results":"results.json"},"results":{"schema":"evopolicygym/leaderboard-results/v1","generated_at":"2026-08-11T03:49:07Z","final_reruns":3,"test_configurations":[{"id":"tetris-strict-128-1-64-v1","label":{"en":"Strict \xb7 128 train / 1 validation / 64 assessment","zh":"\u4e25\u683c\u6a21\u5f0f \xb7 128 \u8bad\u7ec3 / 1 Validation / 64 Assessment"},"description":{"en":"Use all 128 training Episodes, select one Program with 1 Validation Episode, then measure it on 64 held-out Assessment Episodes.","zh":"\u7528\u5b8c 128 \u4e2a\u8bad\u7ec3 Episode\uff0c\u901a\u8fc7 1 \u4e2a Validation Episode \u9009\u62e9\u4e00\u4e2a Program\uff0c\u518d\u5728 64 \u4e2a held-out Assessment Episodes \u4e0a\u8bc4\u6d4b\u3002"},"parameters":{"run_seed":20260810,"training_split":"train","episode_budget":128,"episode_pool_size":128,"max_submissions":128,"max_episodes_per_submission":8,"finish_budget_policy":"require_budget_exhaustion","validation_split":"validation","validation_episodes_per_candidate":1,"validation_max_candidates":1,"assessment_split":"test","assessment_episodes":64,"episode_timeout_seconds":1800,"agent_timeout_seconds":21600}}],"environments":[{"id":"tetris-v5","display":"Atari Tetris","category":"Arcade control","order":0,"primary_metric":"assessment_mean_return","score_direction":"maximize","score_decimal_places":6,"configuration_ids":["tetris-strict-128-1-64-v1"],"default_configuration_id":"tetris-strict-128-1-64-v1","summary":{"en":"Pixel-observation Tetris control in ALE, ranked by mean return across 64 held-out Episodes.","zh":"ALE \u50cf\u7d20\u89c2\u6d4b Tetris \u63a7\u5236\u73af\u5883\uff0c\u6309 64 \u4e2a held-out Episode \u7684 mean return \u6392\u540d\u3002"}}],"entries":[{"id":"gpt_5_6_sol","display":"GPT-5.6 Sol","harness":"Codex","kind":"agent","thinking_effort":"xhigh","scores":{"tetris-v5":{"tetris-strict-128-1-64-v1":400.375}}},{"id":"gpt_5_6_terra","display":"GPT-5.6 Terra","harness":"Codex","kind":"agent","thinking_effort":"xhigh","scores":{"tetris-v5":{"tetris-strict-128-1-64-v1":3.328125}}},{"id":"gpt_5_6_luna","display":"GPT-5.6 Luna","harness":"Codex","kind":"agent","thinking_effort":"xhigh","scores":{"tetris-v5":{"tetris-strict-128-1-64-v1":1.625}}}],"policy_rollouts":[{"entry_id":"gpt_5_6_sol","environment_id":"tetris-v5","configuration_id":"tetris-strict-128-1-64-v1","artifact":"/media/leaderboard/ale-v1/tetris-v5/sol-assessment-episode-000.gif","media_type":"image/gif","camera":"ale-rgb-v1","split":"test","episode_index":0,"score":522},{"entry_id":"gpt_5_6_terra","environment_id":"tetris-v5","configuration_id":"tetris-strict-128-1-64-v1","artifact":"/media/leaderboard/ale-v1/tetris-v5/terra-assessment-episode-000.gif","media_type":"image/gif","camera":"ale-rgb-v1","split":"test","episode_index":0,"score":3},{"entry_id":"gpt_5_6_luna","environment_id":"tetris-v5","configuration_id":"tetris-strict-128-1-64-v1","artifact":"/media/leaderboard/ale-v1/tetris-v5/luna-assessment-episode-000.gif","media_type":"image/gif","camera":"ale-rgb-v1","split":"test","episode_index":0,"score":2}]}}')}}]);

Line numbers count LF bytes from the start of the resource, as the search results do. Vendor segments are library code the classifier recognised; they are stored but not indexed. Bytes are shown as Latin1 characters, one per byte.