World-3 · web and tooling · World-3 collaboration. gemini-3.8-flash on Pi, 26 agents, connected: 746 sessions, 154 agent hours, 5.57 B tokens, $1,350 at list prices. From data/per-release-v3/41.3 in the paper’s figures repository at 9014a8cc, fetched from self-evolving/experiment-41-3 at 58260440 on Hugging Face. The states and releases come from the checkpoint self-evolving/experiment-41-3 @ 0ad42a57 (branch main). The score is rebuilt from the forge’s history, one state every two hours, each task carried at its last measured value.
Scarry 1.305 +30.5% over the seed at the last reading, after 679 sessions −0.8 pp from the previous reading
Sstrict 1.294 a package with a failed check or benchmark scores 1.0 1 package failed at the last reading
Movers +12 / −0 tasks beyond their noise floor at the last reading
Readings 11 latest Sep 24, 23:01 UTC
Run 746 sessions completed work sessions eval set v3.x · speedup over the seed sessions time
+20.0% seed 0 200 400 600 completed sessions Release: aiosignal v1.4.1 Releases: fastapi 0.141.2, referencing v0.37.1, jsonschema-specifications v2026.9.1 Release: iniconfig v2.5.0 Release: fastapi 0.141.3 Release: cloudpickle v3.1.3 Release: rich v15.0.1 Release: h11 v0.17.0 Release: referencing v0.37.2 Release: pydantic v2.13.5 Release: pydantic-settings v2.14.3 Releases: pyyaml 6.0.4, attrs 26.2.0 Release: rich v15.0.2 Release: fastapi 0.141.4 Release: cloudpickle v3.1.4 Releases: h11 v0.18.0, pydantic-settings v2.14.4 Release: pydantic v2.13.6 Release: aiosignal v1.4.2 Release: rich v15.0.3 Release: cloudpickle v3.1.5 Release: referencing v0.37.3 Releases: h11 v0.19.0, pydantic-settings v2.14.5 Release: fastapi 0.141.5 Release: cloudpickle v3.1.6 Release: attrs 26.3.0 Releases: pydantic v2.13.7, pydantic-settings v2.14.6 Release: h11 v0.20.0 Release: referencing v0.37.4 Releases: cloudpickle v3.1.7, rich v15.0.4 Release: pyyaml 6.0.5 Release: fastapi 0.141.6 Release: attrs 26.4.0 Release: pydantic v2.13.8 Release: h11 v0.21.0 Release: aiosignal v1.4.3 Release: cloudpickle v3.1.8 Release: referencing v0.37.5 Release: attrs 26.5.0 Release: rich v15.0.5 Releases: h11 v0.22.0, pydantic v2.13.9 Release: pyyaml 6.0.6 Releases: pydantic-settings v2.14.7, fastapi 0.141.7 Release: jsonschema v4.26.1 Releases: aiosignal v1.4.4, rich v15.0.6 Release: attrs 26.6.0 Release: referencing v0.37.6 Release: h11 v0.23.0 Release: cloudpickle v3.1.9 Release: pydantic-settings v2.14.8 Releases: packaging 26.3, propcache v0.5.7 Release: rich v15.0.7 Release: attrs 26.7.0 Release: fastapi 0.141.8 Release: pydantic v2.13.10 Release: pyyaml 6.0.7 Release: cloudpickle v3.1.10 Releases: attrs 26.8.0, h11 v0.24.0 Release: pyyaml 6.0.8 Release: rich v15.0.8 Releases: cloudpickle v3.1.11, pydantic-settings v2.14.9 Release: h11 v0.25.0 Releases: fastapi 0.141.9, pydantic v2.13.11, aiosignal v1.4.5 Release: attrs 26.9.0 Release: rich v15.0.9 Release: jsonschema v4.26.2 Release: pydantic-settings v2.14.10 Release: h11 v0.26.0 Releases: referencing v0.37.7, cloudpickle v3.1.12 Release: pydantic v2.13.12 Release: attrs 26.10.0 Release: fastapi 0.141.10 Release: h11 v0.27.0 Release: pydantic v2.13.13 Releases: pydantic-settings v2.14.11, cloudpickle v3.1.13, pyyaml 6.0.9 Release: jsonschema v4.26.3 Release: packaging 26.4 Release: aiosignal v1.4.6 Releases: pydantic v2.13.14, h11 v0.28.0, rich v15.0.10 Release: cloudpickle v3.1.14 Release: frozenlist v1.9.9 Release: pyyaml 6.0.10 Releases: referencing v0.37.8, pydantic-settings v2.14.12 Releases: pluggy 1.18.0, aiosignal v1.4.7 Release: fastapi 0.141.11 Release: h11 v0.29.0 Release: cloudpickle v3.1.15 Release: pyyaml 6.0.11 Release: packaging 26.5 Release: h11 v0.30.0 Releases: attrs 26.11.0, cloudpickle v3.1.16, jsonschema v4.26.4 Release: pydantic v2.13.15 Releases: fastapi 0.141.12, pyyaml 6.0.12, pydantic-settings v2.14.13 Release: h11 v0.31.0 Release: cloudpickle v3.1.17 Releases: pyyaml 6.0.13, rich v15.0.11, attrs 26.12.0 Release: pydantic-settings v2.14.14 Release: jsonschema v4.26.5 Release: uvicorn 0.51.7 Release: pyyaml 6.0.14 Release: attrs 26.13.0 Release: cloudpickle v3.1.18 Release: jsonschema v4.26.6 Release: rich v15.0.12 Release: packaging 26.6 Release: pyyaml 6.0.15 Release: jsonschema v4.26.7 Release: rich v15.0.13 Release: pyyaml 6.0.16 Release: aiohttp v3.14.3 Release: pyyaml 6.0.17 Releases: rich v15.0.14, pydantic-settings v2.14.15 Release: rich v15.0.15 Release: pydantic-settings v2.14.16 Releases: rich v15.0.16, pydantic-settings v2.14.17 Release: cloudpickle v3.1.19 Release: h11 v0.32.0 Releases: fastapi 0.141.13, attrs 26.14.0 Release: pydantic v2.13.16 Release: pyyaml 6.0.18 Release: pydantic-settings v2.14.18 Release: cloudpickle v3.1.20 Release: h11 v0.33.0 Release: aiohttp v3.14.4 Releases: attrs 26.15.0, rich v15.0.17 Release: pydantic-settings v2.14.19 Releases: pyyaml 6.0.19, cloudpickle v3.1.21 Release: h11 v0.34.0 Release: attrs 26.16.0 Release: rich v15.0.18 Release: cloudpickle v3.1.22 Release: h11 v0.35.0 Releases: pyyaml 6.0.20, pydantic-settings v2.14.20 Release: attrs 26.17.0 Release: rich v15.0.19 Release: cloudpickle v3.1.23 Release: h11 v0.36.0 Releases: pydantic-settings v2.14.21, pyyaml 6.0.21 Release: attrs 26.18.0 Release: cloudpickle v3.1.24 Release: pydantic-settings v2.14.22 Releases: pyyaml 6.0.22, h11 v0.37.0 Release: fastapi 0.141.14 Release: cloudpickle v3.1.25 Release: attrs 26.19.0 Release: pydantic v2.13.17 Release: jsonschema v4.26.8 Releases: pyyaml 6.0.23, aiohttp v3.14.5 Release: pydantic-settings v2.14.23 Release: h11 v0.38.0 Releases: cloudpickle v3.1.26, attrs 26.20.0, pytest 9.1.22 Releases: pygments 2.20.8, multidict v6.7.3 Releases: tibs v0.5.23, pydantic v2.13.18 Release: pyyaml 6.0.24 Releases: jsonschema v4.26.9, pydantic-settings v2.14.24, h11 v0.39.0 Release: yarl v1.24.17 Releases: rich v15.0.20, pydantic v2.13.19 Release: h11 v0.40.0 Release: attrs 26.21.0 Release: cloudpickle v3.1.27 Release: pydantic-settings v2.14.25 Release: h11 v0.41.0 Release: pydantic v2.13.20 Scarry at each readingSstrict , where it differs a package failed a check or a benchmark 205 releases measuredBreakdown after 679 sessions · Sep 24, 23:01 UTC Scarry 1.305 · Sstrict 1.294 · +31.1% over the tasks with a value · 19/20 tasks measured
openapi-schema-validator · compiled-validation-throughput +320.7% openapi-schema-validator · helper-validation-checked-throughput +80.7% openapi-schema-validator · helper-validation-trusted-throughput +79.2% openapi-schema-validator · first-validation +66.6% kedro · io-bound-runners +51.6% kedro · data-catalog +46.6% kedro · omega-config-loader +38.6% openapi-schema-validator · schema-compilation +28.8% kedro · session-lifecycle +28.2% openapi-spec-validator · validate-synthetic-specs +27.5% openapi-spec-validator · validate-real-world-specs +15.5% dascore · spool-benchmarks +1.2% navigator-api · base-app-handler-construction +0.3% dascore · patch-benchmarks +0.2% dascore · io-benchmarks +0.1% navigator-api · singleton-construction +0.1% tqdm · alternative-progress-libraries 0.0% prefect · cold-imports benchmark failed mitmproxy · proxy-layer-roundtrips 0.0% tqdm · tqdm-loop-overhead −0.5%