19 days ago
-
Replace failed KSEMSTAMP run with validated 250 C4 + 250 BookSum results
Corrected domain assets and margin 0.035; 500 clean outputs, 500 GPT rewrites, 3000 MarkLLM edits, raw scores and empirical ROC. Preserve other methods. Remove obsolete KSEMSTAMP parts and update comparison tables.
21 days ago
-
Add ten-method overview and configuration table
-
Benchmark sync at 2026-09-24 06:29:20 UTC: file benchmark/mpac_config.json
-
Benchmark sync at 2026-09-24 06:29:20 UTC: file benchmark/MPAC_RESULTS_20260924.md
-
Benchmark sync at 2026-09-24 06:29:20 UTC: file benchmark/MPAC_INTEGRATION.md
-
Benchmark sync at 2026-09-24 06:28:20 UTC: file benchmark/MACHINE_SPLIT_STATUS.md
-
Benchmark sync at 2026-09-24 06:28:20 UTC: file benchmark/SEMSTAMP_1GPU_HANDOFF.md
-
Benchmark sync at 2026-09-24 06:28:20 UTC: file benchmark/full_attacked_detection_validation.json
-
Benchmark sync at 2026-09-24 06:28:20 UTC: file benchmark/full_attack_validation.json
-
Benchmark sync at 2026-09-24 06:28:20 UTC: file benchmark/full_clean_validation.json
-
Benchmark sync at 2026-09-24 06:28:20 UTC: file benchmark/compatibility.json
-
Benchmark sync at 2026-09-24 06:28:20 UTC: folder benchmark/markllm_scores
-
Benchmark sync at 2026-09-24 06:28:20 UTC: folder benchmark/markllm_attacks
-
Benchmark sync at 2026-09-24 06:28:20 UTC: folder benchmark/results
-
Benchmark sync at 2026-09-24 06:28:20 UTC: folder benchmark/scores
-
Benchmark sync at 2026-09-24 06:28:20 UTC: folder benchmark/unwatermarked_baseline
-
Benchmark sync at 2026-09-24 06:28:20 UTC: folder benchmark/outputs
-
Benchmark sync at 2026-09-24 06:28:20 UTC: folder benchmark/manifests
-
Sync completed 150-row SEMSTAMP BookSum shard
-
Add validated comparison with partial K-SemStamp
-
Sync stopped K-300 and older methods 1201-1233
-
Sync stopped K-300 and older methods 1101-1200
-
Sync stopped K-300 and older methods 1001-1100
-
Sync stopped K-300 and older methods 901-1000
-
Sync stopped K-300 and older methods 801-900
-
Sync stopped K-300 and older methods 701-800
-
Sync stopped K-300 and older methods 601-700
-
Sync stopped K-300 and older methods 501-600
-
Sync stopped K-300 and older methods 401-500
-
Sync stopped K-300 and older methods 301-400
-
Sync stopped K-300 and older methods 201-300
-
Sync stopped K-300 and older methods 101-200
-
Sync stopped K-300 and older methods 1-100
22 days ago
-
Remove stale C4 result files after sample 13 exclusion
-
Benchmark sync at 2026-09-23 16:52:42 UTC: file benchmark/MACHINE_SPLIT_STATUS.md
-
Benchmark sync at 2026-09-23 16:52:42 UTC: file benchmark/SEMSTAMP_1GPU_HANDOFF.md
-
Benchmark sync at 2026-09-23 16:52:42 UTC: file benchmark/full_attacked_detection_validation.json
-
Benchmark sync at 2026-09-23 16:52:42 UTC: file benchmark/full_attack_validation.json
-
Benchmark sync at 2026-09-23 16:52:42 UTC: file benchmark/full_clean_validation.json
-
Benchmark sync at 2026-09-23 16:52:42 UTC: file benchmark/compatibility.json
-
Benchmark sync at 2026-09-23 16:52:42 UTC: folder benchmark/markllm_scores
-
Benchmark sync at 2026-09-23 16:52:42 UTC: folder benchmark/markllm_attacks
-
Benchmark sync at 2026-09-23 16:52:42 UTC: folder benchmark/scores
-
Benchmark sync at 2026-09-23 16:52:42 UTC: folder benchmark/unwatermarked_baseline
-
Benchmark sync at 2026-09-23 16:52:42 UTC: folder benchmark/outputs
-
Benchmark sync at 2026-09-23 16:52:42 UTC: folder benchmark/manifests
-
Add one-GPU semantic machine handoff
-
Add one-GPU semantic machine handoff
-
Benchmark sync at 2026-09-23 16:32:21 UTC: folder benchmark/markllm_scores
-
Benchmark sync at 2026-09-23 16:32:21 UTC: folder benchmark/markllm_attacks