AniketAsla commited on
Commit
ee7ca75
·
verified ·
1 Parent(s): 4b93822

deploy: update reports/eval_report.md

Browse files
Files changed (1) hide show
  1. reports/eval_report.md +38 -38
reports/eval_report.md CHANGED
@@ -1,38 +1,38 @@
1
- # Evaluation Report
2
-
3
- Generated at: 2026-04-25T18:12:09.069260+00:00
4
- Base URL: http://localhost:7860
5
- Tasks: clean_claim, contradictory_claim, coordinated_fraud, distribution_shift_claim, identity_fraud
6
- Seeds: 7, 11, 13, 19, 25
7
- Distinct variant_ids: [0, 1, 2, 3, 4]
8
-
9
- | Task | Seed | Variant | Steps | Done | Reward | Evidence Quality | Exploit Penalty |
10
- |---|---:|---:|---:|:---:|---:|---:|---:|
11
- | clean_claim | 7 | 2 | 4 | yes | 0.8725 | 1.0000 | 0.0000 |
12
- | clean_claim | 11 | 1 | 4 | yes | 0.8725 | 1.0000 | 0.0000 |
13
- | clean_claim | 13 | 3 | 4 | yes | 0.8725 | 1.0000 | 0.0000 |
14
- | clean_claim | 19 | 4 | 4 | yes | 0.8725 | 1.0000 | 0.0000 |
15
- | clean_claim | 25 | 0 | 4 | yes | 0.8725 | 1.0000 | 0.0000 |
16
- | contradictory_claim | 7 | 2 | 8 | yes | 0.7497 | 1.0000 | 0.0000 |
17
- | contradictory_claim | 11 | 1 | 8 | yes | 0.7497 | 1.0000 | 0.0000 |
18
- | contradictory_claim | 13 | 3 | 8 | yes | 0.7497 | 1.0000 | 0.0000 |
19
- | contradictory_claim | 19 | 4 | 8 | yes | 0.7497 | 1.0000 | 0.0000 |
20
- | contradictory_claim | 25 | 0 | 8 | yes | 0.7497 | 1.0000 | 0.0000 |
21
- | coordinated_fraud | 7 | 2 | 12 | yes | 0.8230 | 1.0000 | 0.0000 |
22
- | coordinated_fraud | 11 | 1 | 12 | yes | 0.8230 | 1.0000 | 0.0000 |
23
- | coordinated_fraud | 13 | 3 | 12 | yes | 0.8230 | 1.0000 | 0.0000 |
24
- | coordinated_fraud | 19 | 4 | 12 | yes | 0.8230 | 1.0000 | 0.0000 |
25
- | coordinated_fraud | 25 | 0 | 12 | yes | 0.8230 | 1.0000 | 0.0000 |
26
- | distribution_shift_claim | 7 | 2 | 12 | yes | 0.7827 | 1.0000 | 0.0000 |
27
- | distribution_shift_claim | 11 | 1 | 12 | yes | 0.7827 | 1.0000 | 0.0000 |
28
- | distribution_shift_claim | 13 | 3 | 12 | yes | 0.7827 | 1.0000 | 0.0000 |
29
- | distribution_shift_claim | 19 | 4 | 12 | yes | 0.7827 | 1.0000 | 0.0000 |
30
- | distribution_shift_claim | 25 | 0 | 12 | yes | 0.7827 | 1.0000 | 0.0000 |
31
- | identity_fraud | 7 | 2 | 10 | yes | 0.8180 | 1.0000 | 0.0000 |
32
- | identity_fraud | 11 | 1 | 10 | yes | 0.8180 | 1.0000 | 0.0000 |
33
- | identity_fraud | 13 | 3 | 10 | yes | 0.8180 | 1.0000 | 0.0000 |
34
- | identity_fraud | 19 | 4 | 10 | yes | 0.8180 | 1.0000 | 0.0000 |
35
- | identity_fraud | 25 | 0 | 10 | yes | 0.8180 | 1.0000 | 0.0000 |
36
-
37
- Average Reward: 0.8092
38
- Completion Rate: 100.00%
 
1
+ # Evaluation Report
2
+
3
+ Generated at: 2026-04-25T18:12:09.069260+00:00
4
+ Base URL: http://localhost:7860
5
+ Tasks: clean_claim, contradictory_claim, coordinated_fraud, distribution_shift_claim, identity_fraud
6
+ Seeds: 7, 11, 13, 19, 25
7
+ Distinct variant_ids: [0, 1, 2, 3, 4]
8
+
9
+ | Task | Seed | Variant | Steps | Done | Reward | Evidence Quality | Exploit Penalty |
10
+ |---|---:|---:|---:|:---:|---:|---:|---:|
11
+ | clean_claim | 7 | 2 | 4 | yes | 0.8725 | 1.0000 | 0.0000 |
12
+ | clean_claim | 11 | 1 | 4 | yes | 0.8725 | 1.0000 | 0.0000 |
13
+ | clean_claim | 13 | 3 | 4 | yes | 0.8725 | 1.0000 | 0.0000 |
14
+ | clean_claim | 19 | 4 | 4 | yes | 0.8725 | 1.0000 | 0.0000 |
15
+ | clean_claim | 25 | 0 | 4 | yes | 0.8725 | 1.0000 | 0.0000 |
16
+ | contradictory_claim | 7 | 2 | 8 | yes | 0.7497 | 1.0000 | 0.0000 |
17
+ | contradictory_claim | 11 | 1 | 8 | yes | 0.7497 | 1.0000 | 0.0000 |
18
+ | contradictory_claim | 13 | 3 | 8 | yes | 0.7497 | 1.0000 | 0.0000 |
19
+ | contradictory_claim | 19 | 4 | 8 | yes | 0.7497 | 1.0000 | 0.0000 |
20
+ | contradictory_claim | 25 | 0 | 8 | yes | 0.7497 | 1.0000 | 0.0000 |
21
+ | coordinated_fraud | 7 | 2 | 12 | yes | 0.8230 | 1.0000 | 0.0000 |
22
+ | coordinated_fraud | 11 | 1 | 12 | yes | 0.8230 | 1.0000 | 0.0000 |
23
+ | coordinated_fraud | 13 | 3 | 12 | yes | 0.8230 | 1.0000 | 0.0000 |
24
+ | coordinated_fraud | 19 | 4 | 12 | yes | 0.8230 | 1.0000 | 0.0000 |
25
+ | coordinated_fraud | 25 | 0 | 12 | yes | 0.8230 | 1.0000 | 0.0000 |
26
+ | distribution_shift_claim | 7 | 2 | 12 | yes | 0.7827 | 1.0000 | 0.0000 |
27
+ | distribution_shift_claim | 11 | 1 | 12 | yes | 0.7827 | 1.0000 | 0.0000 |
28
+ | distribution_shift_claim | 13 | 3 | 12 | yes | 0.7827 | 1.0000 | 0.0000 |
29
+ | distribution_shift_claim | 19 | 4 | 12 | yes | 0.7827 | 1.0000 | 0.0000 |
30
+ | distribution_shift_claim | 25 | 0 | 12 | yes | 0.7827 | 1.0000 | 0.0000 |
31
+ | identity_fraud | 7 | 2 | 10 | yes | 0.8180 | 1.0000 | 0.0000 |
32
+ | identity_fraud | 11 | 1 | 10 | yes | 0.8180 | 1.0000 | 0.0000 |
33
+ | identity_fraud | 13 | 3 | 10 | yes | 0.8180 | 1.0000 | 0.0000 |
34
+ | identity_fraud | 19 | 4 | 10 | yes | 0.8180 | 1.0000 | 0.0000 |
35
+ | identity_fraud | 25 | 0 | 10 | yes | 0.8180 | 1.0000 | 0.0000 |
36
+
37
+ Average Reward: 0.8092
38
+ Completion Rate: 100.00%