Clarify dataset documentation and repair viewer configuration

MANIFEST.sha256

@@ -1,4 +1,4 @@
-2776f20527cd2f153fbbbf3f1f5ad2a49f03052ffc3cc7609e9b96c3ee5e832f  README.md
+db0bb1450c3b3e96a28a5de6cb320ba885ede65ad799520231f1aad5cf2522e9  README.md
 0f9190c2f4c7e438469929c100f809423efbd6394c71758cb3046fcfed45e793  WITHHELD.json
 95d33520836ca41f96c3ebb5bb1facc225a8e61c9f74143fadd9443e1b13bff4  annotations/blind_variation/CODEBOOK_VARIATION.md
 280dffce7e5de8d042d7c82264e2511b9943b84160eb00e5b881d7b3a193363c  annotations/blind_variation/README.md
@@ -24,7 +24,7 @@ a7549f6bd7782988a657f1fe8dc54254ab00f8b12753464dbc5f1e51a20766c3  board/SENSITIV
 a038cbbe66b1dd489686b9514bf98e3c4f7f199c03409d8260359e4e54f6d76b  board/SUPPORTED.jsonl
 0ad79b34e45ce150d2a0a89a7cc2375ddca4dea9a6a1a4602d9e1cf87806ee11  board/WORLDS.jsonl
 97cebb0dfb3d13aeea7781021c116e711089c620b1eb20f32309795ef212f8b1  board/report.log
-7e26210d57b8016f7a5137ee25f73454cafca4b00df63a20d66b182238bc55f9  croissant.json
+56b9b8ce9500306dcdacb8f1feac4d33efb3fd4ccafa0402edb7e32201b5e6fb  croissant.json
 000329955b6bd7e119e55007a648c3fe32928c7de76351c1d502cf3b41b311b1  environment/base-investigation/campaign/blobs/000329955b6bd7e119e55007a648c3fe32928c7de76351c1d502cf3b41b311b1
 00e8c824e4744d06d5612fb819aeab06f285f12604ab2c523dc4fbfe879cdec2  environment/base-investigation/campaign/blobs/00e8c824e4744d06d5612fb819aeab06f285f12604ab2c523dc4fbfe879cdec2
 028331253510ebc1155ff317c93843124d0ab9ee8e87a1c56fe2ee17e97e2bf2  environment/base-investigation/campaign/blobs/028331253510ebc1155ff317c93843124d0ab9ee8e87a1c56fe2ee17e97e2bf2

README.md

@@ -15,8 +15,6 @@ configs:
 - config_name: investigations
   data_files: investigations.parquet
   default: true
-- config_name: worlds
-  data_files: worlds/worlds.jsonl
 - config_name: routes
   data_files: worlds/routes.parquet
 ---
@@ -32,7 +30,7 @@ The [research post](https://dynamicalsystems.ai/blog/sdl-2) explains the finding
 ## Start here
 
 - Use `investigations.parquet` to compare decisions, scores and experiment spending, and to locate each trace.
-- Use `worlds/routes.parquet` to inspect what each processing route can reliably deliver.
+- Use `worlds/worlds.jsonl` to inspect each world and `worlds/routes.parquet` for its route outcomes.
 - Follow the [reproduction instructions](https://github.com/Dynamical-Systems-Research/sdl2#reproduce-the-results) to regenerate the published results from the released records.
 
 ## Contents
@@ -76,9 +74,11 @@ The [grading rules](https://github.com/Dynamical-Systems-Research/sdl2#scoring)
 
 The evaluation used 32 simulated worlds, balanced between 16 with a working route and 16 without one. Worlds were drawn from the supplied physics model’s uncertainty and selected on hidden properties, with quotas fixed before any agent ran. Language models used the same Pi 1.4.0 harness, at high reasoning effort, offline and text-only. Each model was pinned to one provider through OpenRouter.
 
+The `wall_seconds` column in `investigations.parquet` records elapsed time from the original run summary. In `board/FEATURES.jsonl`, it uses summed event spans where available, excluding gaps between attempts, and otherwise retains the summary value.
+
 The study ran on Linux aarch64. The release’s grading and reproduction commands have also been verified on Linux x86_64. Running a new investigation requires the released aarch64 container images. Platform requirements and evaluation issues are documented in the [repository README](https://github.com/Dynamical-Systems-Research/sdl2).
 
-The physical record under `environment/` comes from our experimental campaign with Rensselaer Polytechnic Institute (RPI). RPI cleared its data for this release. Its correspondence is withheld.
+The experimental record under `environment/` comes from our experimental campaign with Rensselaer Polytechnic Institute (RPI). RPI cleared its data for this release. Its correspondence is withheld.
 
 </details>
 
@@ -109,7 +109,7 @@ Assistant messages, reasoning and tool calls in the 192 language-model traces ar
 
 ## Uses and limits
 
-Use this dataset to evaluate agents on SDL-2, study their investigations and check the published findings. The evaluation measurements are simulated. The twin’s density model is calibrated on the experimental campaign and its magnetic model on earlier samples. The physical records used for calibration are included separately under `environment/`.
+Use this dataset to evaluate agents on SDL-2, study their investigations and check the published findings. The evaluation measurements are simulated. The twin’s density model is calibrated on the experimental campaign and its magnetic model on earlier samples. The experimental records used for calibration are included separately under `environment/`.
 
 ## License and citation
 

croissant.json

@@ -49,7 +49,7 @@
  "@type": "sc:Dataset",
  "conformsTo": "http://mlcommons.org/croissant/1.0",
  "name": "SDL-2",
- "description": "SDL-2: 32 simulated NdFeB magnet powders with their hidden parameters and the true outcome of all 8,064 processing routes in each, and 256 complete investigations by six language models and two classical experiment-design methods, with their Outcome, Supported and Both scores.",
+ "description": "SDL-2 contains 32 simulated NdFeB magnet powders with their hidden parameters and the true outcome of all 8,064 processing routes in each, and 256 investigations by six language models and two classical experiment-design methods, with their Outcome, Supported and Both scores.",
  "license": "https://creativecommons.org/licenses/by/4.0/",
  "url": "https://huggingface.co/datasets/Dynamical-Systems/sdl2",
  "citeAs": "Barnes, J. (2026). Teaching AI to Explore the Physical World. Dynamical Systems. https://dynamicalsystems.ai/blog/sdl-2",