SAVRN
Search Contact SAVRN

Dataset · Text generation

nemotron-student-fail-v41-clean-thinking

by Zhiyuan Hu zhiyuanhucs/nemotron-student-fail-v41-clean-thinking

DeepSeek-V4.1 reward-1 trajectories for tasks on which the Nemotron student did not obtain reward 1. This release was rebuilt from the complete reward-1 audit under v56-shell-quote-aware-package-probes+v57-target-source-recheck.

Rows—
Configurations—
Size470.5 MB
Licenseother
AccessPublicly accessible
Monthly Downloads8.3k

Dataset Card

DeepSeek-V4.1 reward-1 trajectories for tasks on which the Nemotron student did not obtain reward 1. This release was rebuilt from the complete reward-1 audit under v56-shell-quote-aware-package-probes+v57-target-source-recheck. 每个目录内的文件名都直接说明:是一题一条还是全部 trials、是否为 audit、thinking/action 是否干净。 本版重新检查了每条候选轨迹的实际 tool action。查询、下载或读取当前任务项目的已发布源码时, 即使 verifier 给出 reward=1,也不会进入上述训练文件。审计隔离清单位于 data/03allreward1trials/targetpublishedsourcelookupquarantineaudit.jsonl.gz 167 条 trial)。 data/04reward1successfulbuthackingfailed/ 也单独排除了 SWE-bench Pro Verified is exported under a separate benchmark path: The Pro collector continuously rerolls unresolved tasks. The current export has 138 verifier-correct…

Excerpt from the card by Zhiyuan Hu, licensed other.

Details

Repository
zhiyuanhucs/nemotron-student-fail-v41-clean-thinking
Publisher
Zhiyuan Hu
Task category
Text generation
Tags
Not stated by the source
Size category
Not stated by the source
Languages
en
Revision
7d4038616e7280af7c08e3d281191250a8354c87
Last updated
2026-09-29

Files

88 files, 470.5 MB in total.

Data86 files · 470.5 MB
Documentation1 file · 12.5 KB
Repository1 file · 2.5 KB
Every file
FileTypeSizeSHA-256
blocked_attempt_control_manifest.jsonlData30.2 KB—
data/01_clean_thinking_and_actions/all_successful_trials.jsonl.gzData1.1 MB5a95cc2b4614
data/01_clean_thinking_and_actions/one_trajectory_per_task.jsonl.gzData893.5 KB321aede00557
data/02_clean_actions_dirty_thinking_removed/all_successful_trials.jsonl.gzData37.7 MB1eab7ca8f445
data/02_clean_actions_dirty_thinking_removed/one_trajectory_per_task.jsonl.gzData13.6 MBdc6ecb3e9a23
data/03_all_reward1_trials/all_trials__audit_index.jsonl.gzData47.5 KB514af1c17f95
data/03_all_reward1_trials/clean_actions_dirty_thinking_removed__all_trials.jsonl.gzData37.7 MB1eab7ca8f445
data/03_all_reward1_trials/clean_thinking_and_actions__all_trials.jsonl.gzData1.1 MB5a95cc2b4614
data/03_all_reward1_trials/raw_original_trajectories_before_thinking_redaction__action_clean_reward1.jsonl.gzData38.4 MBbaec4c8defa1
data/03_all_reward1_trials/rejected_trials__audit_only.jsonl.gzData33.5 KB1fba9247b43f
data/03_all_reward1_trials/target_published_source_lookup__quarantine_audit.jsonl.gzData44.9 KB8b4a80ab33ee
data/04_reward1_successful_but_hacking_failed/all_guard_rejected_dirty_generations.jsonl.gzData1.3 MB93e548f73d3a
data/04_reward1_successful_but_hacking_failed/all_successful_trials.jsonl.gzData29.6 MBd07ed37086c0
data/04_reward1_successful_but_hacking_failed/all_trials__audit_index.jsonl.gzData15.9 KB845d70908852
data/04_reward1_successful_but_hacking_failed/one_trajectory_per_task.jsonl.gzData14.1 MBf93398e6f472
data/04_reward1_successful_but_hacking_failed/strong_proof_guard_blocked_before_execution/all_successful_trials.jsonl.gzData14.2 MBbb6f27e4ddc8
data/04_reward1_successful_but_hacking_failed/strong_proof_guard_blocked_before_execution/one_trajectory_per_task.jsonl.gzData5.5 MB17afc41798a7
data/04_reward1_successful_but_hacking_failed/target_published_source_lookup__quarantine_audit.jsonl.gzData7.7 KB856972f20c53
data/04_reward1_successful_but_hacking_failed/weak_proof_no_successful_answer_access_detected/all_successful_trials.jsonl.gzData15.4 MBb6a75fe46db3
data/04_reward1_successful_but_hacking_failed/weak_proof_no_successful_answer_access_detected/one_trajectory_per_task.jsonl.gzData8.5 MB3f42fffbede7
data/05_offline_no_guard_experiment/deepseek_v4_1/all_reward1_without_successful_hack.jsonl.gzData53 B27fe8834bfd5
data/05_offline_no_guard_experiment/deepseek_v4_1/all_scored_tasks__audit.jsonl.gzData5.8 KBdbb5606333b5
data/05_offline_no_guard_experiment/deepseek_v4_1/attempted_test_changes_but_external_verifier_still_passed.jsonl.gzData62 B80937dea99ce
data/05_offline_no_guard_experiment/deepseek_v4_1/clean_thinking_and_actions__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/deepseek_v4_1/dirty_actions_only__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/deepseek_v4_1/dirty_thinking_and_dirty_actions__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/deepseek_v4_1/dirty_thinking_only__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/deepseek_v4_1/legacy_public_egress_untrusted_reward1.jsonl.gzData4.9 MB631e76807426
data/05_offline_no_guard_experiment/deepseek_v4_1_full_pool/all_reward1_without_successful_hack.jsonl.gzData15.0 MB1dad8ed14165
data/05_offline_no_guard_experiment/deepseek_v4_1_full_pool/all_scored_tasks__audit.jsonl.gzData78.2 KB1cc3747dda6d
data/05_offline_no_guard_experiment/deepseek_v4_1_full_pool/attempted_test_changes_but_external_verifier_still_passed.jsonl.gzData2.2 MBacb7cb0d313e
data/05_offline_no_guard_experiment/deepseek_v4_1_full_pool/clean_thinking_and_actions__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/deepseek_v4_1_full_pool/dirty_actions_only__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/deepseek_v4_1_full_pool/dirty_thinking_and_dirty_actions__reward1.jsonl.gzData11.1 MBf3e6ba40344d
data/05_offline_no_guard_experiment/deepseek_v4_1_full_pool/dirty_thinking_only__reward1.jsonl.gzData3.8 MB759787ecf43a
data/05_offline_no_guard_experiment/deepseek_v4_1_full_pool/legacy_public_egress_untrusted_reward1.jsonl.gzData7.4 MB55ad09c8b573
data/05_offline_no_guard_experiment/deepseek_v4_1_low_reasoning_pilot/all_reward1_without_successful_hack.jsonl.gzData53 B27fe8834bfd5
data/05_offline_no_guard_experiment/deepseek_v4_1_low_reasoning_pilot/all_scored_tasks__audit.jsonl.gzData280 B66ee847e8eb8
data/05_offline_no_guard_experiment/deepseek_v4_1_low_reasoning_pilot/clean_thinking_and_actions__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/deepseek_v4_1_low_reasoning_pilot/dirty_actions_only__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/deepseek_v4_1_low_reasoning_pilot/dirty_thinking_and_dirty_actions__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/deepseek_v4_1_low_reasoning_pilot/dirty_thinking_only__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/deepseek_v4_1_prompt_only_pilot/all_reward1_without_successful_hack.jsonl.gzData91.5 KB832c56e62d63
data/05_offline_no_guard_experiment/deepseek_v4_1_prompt_only_pilot/all_scored_tasks__audit.jsonl.gzData472 B6c26c10f061d
data/05_offline_no_guard_experiment/deepseek_v4_1_prompt_only_pilot/clean_thinking_and_actions__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/deepseek_v4_1_prompt_only_pilot/dirty_actions_only__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/deepseek_v4_1_prompt_only_pilot/dirty_thinking_and_dirty_actions__reward1.jsonl.gzData91.5 KBb31c1a79caed
data/05_offline_no_guard_experiment/deepseek_v4_1_prompt_only_pilot/dirty_thinking_only__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/deepseek_v4_1_vs_nemotron_3_ultra/all_paired_tasks__audit.jsonl.gzData742 Bcf5c058d8d6c
data/05_offline_no_guard_experiment/deepseek_v4_1_vs_nemotron_3_ultra/v4_1_success_nemotron_failed.jsonl.gzData57 Ba47124ddadc5
data/05_offline_no_guard_experiment/nemotron_3_ultra/all_reward1_without_successful_hack.jsonl.gzData53 B27fe8834bfd5
data/05_offline_no_guard_experiment/nemotron_3_ultra/all_scored_tasks__audit.jsonl.gzData5.9 KBba27b8283677
data/05_offline_no_guard_experiment/nemotron_3_ultra/attempted_test_changes_but_external_verifier_still_passed.jsonl.gzData62 B80937dea99ce
data/05_offline_no_guard_experiment/nemotron_3_ultra/clean_thinking_and_actions__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/nemotron_3_ultra/dirty_actions_only__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/nemotron_3_ultra/dirty_thinking_and_dirty_actions__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/nemotron_3_ultra/dirty_thinking_only__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/nemotron_3_ultra/legacy_public_egress_untrusted_reward1.jsonl.gzData1.0 MBff71ecf49df8
data/05_offline_no_guard_experiment/nemotron_3_ultra_step256_retry/all_reward1_without_successful_hack.jsonl.gzData53 B27fe8834bfd5
data/05_offline_no_guard_experiment/nemotron_3_ultra_step256_retry/all_scored_tasks__audit.jsonl.gzData277 Ba309005b0cef
data/05_offline_no_guard_experiment/nemotron_3_ultra_step256_retry/attempted_test_changes_but_external_verifier_still_passed.jsonl.gzData62 B80937dea99ce
data/05_offline_no_guard_experiment/nemotron_3_ultra_step256_retry/clean_thinking_and_actions__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/nemotron_3_ultra_step256_retry/dirty_actions_only__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/nemotron_3_ultra_step256_retry/dirty_thinking_and_dirty_actions__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/nemotron_3_ultra_step256_retry/dirty_thinking_only__reward1.jsonl.gzData32 B9022aad581c6
data/05_offline_no_guard_experiment/nemotron_3_ultra_step256_retry/legacy_public_egress_untrusted_reward1.jsonl.gzData65 Ba064e420ef57
data/06_swebench_pro_verified/all_reward1_trials__audit_index.jsonl.gzData202.2 KB67c36d4097ef
data/06_swebench_pro_verified/clean_actions_dirty_thinking_removed__all_trials.jsonl.gzData34.4 MBa2b575529960
data/06_swebench_pro_verified/clean_actions_dirty_thinking_removed__one_per_task.jsonl.gzData30.7 MB31187b1a1098
data/06_swebench_pro_verified/clean_thinking_and_actions__all_trials.jsonl.gzData40 Beec3ad3a60e1
data/06_swebench_pro_verified/clean_thinking_and_actions__one_per_task.jsonl.gzData32 B9022aad581c6
data/07_official_swebench_pro_verified_baseline/deepseek_v4_1/all_valid_once_per_task.jsonl.gzData69.8 MB531617460705
data/07_official_swebench_pro_verified_baseline/deepseek_v4_1/infrastructure_failures_excluded.jsonl.gzData5.6 KB653094249f4b
data/07_official_swebench_pro_verified_baseline/deepseek_v4_1/resolved_once_per_task.jsonl.gzData27.2 MBe38e1fdfcf06
data/07_official_swebench_pro_verified_baseline/nemotron_3_ultra/all_valid_once_per_task.jsonl.gzData30.5 MB7a7bf425c702
data/07_official_swebench_pro_verified_baseline/nemotron_3_ultra/infrastructure_failures_excluded.jsonl.gzData2.5 KBdd3cc2a3ddcd
data/07_official_swebench_pro_verified_baseline/nemotron_3_ultra/resolved_once_per_task.jsonl.gzData11.8 MBc10ed24426fc
data/08_source_link_sanitized_recovery/clean_actions_dirty_thinking_removed__one_per_task.jsonl.gzData325.5 KB4392c77e088d
data/08_source_link_sanitized_recovery/clean_thinking_and_actions__one_per_task.jsonl.gzData67 B1c2d16c635ea
isolated_no_guard_manifest.jsonlData38.4 KB—
manifest.jsonlData95.5 KB—
official_swebench_pro_verified_baseline_manifest.jsonlData299.1 KB—
rejected-summary.jsonData583 B—
rejected.jsonlData59.9 KB—
summary.jsonData14.3 KB—
swebench_pro_verified_manifest.jsonlData73.7 KB—
README.mdDocumentation12.5 KB—
.gitattributesRepository2.5 KB—

License and Download

License
other
Access
No access gate
Download from Zhiyuan Hu

Released by Zhiyuan Hu through its official repository on Hugging Face.