Dataset Viewer
The dataset viewer is not available for this subset.
Cannot get the split names for the config 'default' of the dataset.
Exception:    SplitsNotFoundError
Message:      The split names could not be parsed from the dataset config.
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 286, in get_dataset_config_info
                  for split_generator in builder._split_generators(
                                         ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      StreamingDownloadManager(base_path=builder.base_path, download_config=download_config)
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/harbor/harbor.py", line 171, in _split_generators
                  raise DataFilesNotFoundError("No task.toml or instruction.md files found")
              datasets.exceptions.DataFilesNotFoundError: No task.toml or instruction.md files found
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/split_names.py", line 68, in compute_split_names_from_streaming_response
                  for split in get_dataset_split_names(
                               ~~~~~~~~~~~~~~~~~~~~~~~^
                      path=dataset,
                      ^^^^^^^^^^^^^
                      config_name=config,
                      ^^^^^^^^^^^^^^^^^^^
                      token=hf_token,
                      ^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 340, in get_dataset_split_names
                  info = get_dataset_config_info(
                      path,
                  ...<6 lines>...
                      **config_kwargs,
                  )
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 291, in get_dataset_config_info
                  raise SplitsNotFoundError("The split names could not be parsed from the dataset config.") from err
              datasets.inspect.SplitsNotFoundError: The split names could not be parsed from the dataset config.

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

[refusal features 10f] (session c7c7e46b-a324-4eef-ae90-42ed2c011745, agent refusal-features-10f, 2026-10-30)

Transcripts of the refusal-feature search on the 10f transcoder adapter

Every greedy generation of experiments/refusal_tracing/writeup.md §15 (repo Sid-MB/transcoder-adapters), one markdown file per prompt with the request, the adapter's 256-token answer under each condition and the Qwen3-8B judge's label, confidence and rationale. The prompts are HarmBench requests (harmful), the refusal ladder's 30 benign instructions, and the 24 verified GCG requests plain and suffixed; the adapter is siddharthmb/2026.TA.gemma2_2b_headline_2026-10f_tc8192_decb_l1w0.0003_auxk512_ctr_fp32_tarbb_lb2_sl17879940. Each folder is the transcripts/ dir of one run under experiments/refusal_tracing/results/refusal_features_10f/, whose results.json, summary.md and run_manifest.json (the full command) sit beside the breadcrumb transcripts.txt.

folder script and flags Slurm job log
activations/ refusal_feature_activations.py --prompt_sets harmful benign harmbench gcg_plain gcg_suffixed (341 prompts, no steering) 18202064 logs/refusal_features/20261030_160023_18202064.{out,err}
steering/ steer_refusal_features.py (141 prompts; baseline, ablate, amplify x0.5, x1, x2 of the 79 selected features) 18202065 logs/refusal_features/20261030_160023_18202065.{out,err}
steering_gentle/ steer_refusal_features.py --conditions baseline amplify:0.05 amplify:0.1 amplify:0.2 18202398 logs/refusal_features/20261030_162735_18202398.{out,err}
steering_separators/ steer_refusal_features.py --separators_from .../activations/results.json --min_auroc 0.7 --conditions baseline ablate amplify:1 amplify:2 amplify:4 amplify:8 (6 features) 18202399 logs/refusal_features/20261030_162735_18202399.{out,err}
steering_separators_low/ the same six at --conditions baseline amplify:0.1 amplify:0.25 amplify:0.5 18202805 logs/refusal_features/20261030_165935_18202805.{out,err}

Restore a folder into the repo with uv run hf download <this repo> --repo-type dataset --include "<folder>/*" --local-dir /tmp/hf_restore && mv /tmp/hf_restore/<folder> experiments/refusal_tracing/results/refusal_features_10f/<folder>/transcripts.

Downloads last month
588