nltk-subword-tokenization

Task gallery / DeepSWE (long-horizon SWE) / Version 1

Reference solution (oracle): Unverified
No-op resistance check: Unverified
QA verdict: Not recorded
Aggregate eligibility: Eligible version

The task

Add a standard-library-only package nltk/tokenize/subword/ (no new dependency, not even re/regex) with BPE, WordPiece and Unigram; change no existing NLTK module; all names below import from nltk.tokenize.subword. Match the HuggingFace tokenizers Python API (signatures, defaults, behavior) where a counterpart exists; deviations/extras: normalizers (NFC/NFD/NFKC/Lowercase/StripAccents/Strip/Replace/Prepend/BertNormal…

Excerpt only; not a complete executable task specification. Task ID task_3fbbd1e5aaef4a84 · version 1. Selection status: not fully recorded.

Results by model and effort

This exact task version only. Every rate is passes / valid runs with a Wilson 95% interval; n≤2 remains counts only. Macro aggregation does not apply within a single task. Baselines are summarized in the quality gates, not ranked. A passed no-op resistance check means that doing nothing did not solve the task.

Current evidence

Model / effort	Verifier result [95% Wilson]	Diagnostic labels / valid	HARNESS_ERROR · excluded	Outcome classes

Legacy evidence

Model / effort	Verifier result [95% Wilson]	Diagnostic labels / valid	HARNESS_ERROR · excluded	Outcome classes
Claude Haiku 4.5 · default (unverified)	0/0	0 / 0	5	GOOD_FAILURE: 0 BAD_FAILURE: 0 GOOD_SUCCESS: 0 BAD_SUCCESS: 0

Every recorded run

5 recorded trials, including controls. Diagnostic labels do not replace verifier reward. Labels and quality checks are source declarations, not independently attested here. Internal evidence remains private; a missing public URL is shown explicitly. Records with invalid or absent rewards are not scored.

Run / model	Date / source	Reward / eligibility	Diagnostic class / source	Resources	Evidence
`trial_147d8889d8ed4acd` Claude Haiku 4.5 · default (unverified)	Sep 5, 2026 console	0 · excludedharness_error	HARNESS_ERROR frozen-classifier	Tokens: 0 Turns: 1 Cost: $0.0000	Not publicly available
`trial_215162fd5f7a4009` Claude Haiku 4.5 · default (unverified)	Sep 5, 2026 console	0 · excludedharness_error	HARNESS_ERROR frozen-classifier	Tokens: 0 Turns: 1 Cost: $0.0000	Not publicly available
`trial_861b3f7f56f04cc3` Claude Haiku 4.5 · default (unverified)	Sep 5, 2026 console	0 · excludedharness_error	HARNESS_ERROR frozen-classifier	Tokens: 0 Turns: 1 Cost: $0.0000	Not publicly available
`trial_a11bf0685f244fa6` Claude Haiku 4.5 · default (unverified)	Sep 5, 2026 console	0 · excludedharness_error	HARNESS_ERROR frozen-classifier	Tokens: 0 Turns: 1 Cost: $0.0000	Not publicly available
`trial_e0bd9b7ba7d9402e` Claude Haiku 4.5 · default (unverified)	Sep 5, 2026 console	0 · excludedharness_error	HARNESS_ERROR frozen-classifier	Tokens: 0 Turns: 1 Cost: $0.0000	Not publicly available