Fix: scope title to thinking models, add Sept V4 Pro Pattern C data
Browse files- README.md +1 -1
- results/qwen_comparison.csv +2 -2
README.md
CHANGED
|
@@ -133,7 +133,7 @@ Provider pinning is essential. See [LessWrong](https://www.lesswrong.com/posts/K
|
|
| 133 |
|
| 134 |
```bibtex
|
| 135 |
@misc{kumar2026compliance_cot,
|
| 136 |
-
title={Compliance-Induced Fabrication Is Transparent: A Model Forensics Investigation},
|
| 137 |
author={Kumar, Rahul},
|
| 138 |
year={2026},
|
| 139 |
note={Extends arXiv:2605.02398}
|
|
|
|
| 133 |
|
| 134 |
```bibtex
|
| 135 |
@misc{kumar2026compliance_cot,
|
| 136 |
+
title={Compliance-Induced Fabrication Is Transparent in Thinking Models: A Model Forensics Investigation},
|
| 137 |
author={Kumar, Rahul},
|
| 138 |
year={2026},
|
| 139 |
note={Extends arXiv:2605.02398}
|
results/qwen_comparison.csv
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
metric,v4_pro_paper_apr,v4_pro_sept3,qwen_sept3
|
| 2 |
fabrication_rate_ebd_unanswerable,0.843,0.365,0.258
|
| 3 |
-
pattern_c_pct,1.000,,1.000
|
| 4 |
-
pattern_c_n,"129/129",,"41/41"
|
| 5 |
hedged_output_pct,0.142,,0.585
|
| 6 |
avg_fabrication_reasoning_chars,2999,2983,13766
|
| 7 |
avg_refusal_reasoning_chars,828,1336,3607
|
|
|
|
| 1 |
metric,v4_pro_paper_apr,v4_pro_sept3,qwen_sept3
|
| 2 |
fabrication_rate_ebd_unanswerable,0.843,0.365,0.258
|
| 3 |
+
pattern_c_pct,1.000,1.000,1.000
|
| 4 |
+
pattern_c_n,"129/129","58/58","41/41"
|
| 5 |
hedged_output_pct,0.142,,0.585
|
| 6 |
avg_fabrication_reasoning_chars,2999,2983,13766
|
| 7 |
avg_refusal_reasoning_chars,828,1336,3607
|