Spaces:
Running
Running
ratingtesting commited on
Commit ·
ae24bc4
1
Parent(s): dc96594
Update KDS leaderboard: kimi-k3 STRONG (25), Ling re-run (22), Nemotron-3-super PARTIAL
Browse files
README.md
CHANGED
|
@@ -13,7 +13,25 @@ pinned: false
|
|
| 13 |
|
| 14 |
Catches SQL injection, hardcoded secrets, slopsquatting, reward hacking, doom loops, and 23 more failure modes.
|
| 15 |
|
| 16 |
-
**Keelwright Score (KDS) up to 83/100** — proven by adversarial A/B testing.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
|
| 18 |
## Links
|
| 19 |
|
|
|
|
| 13 |
|
| 14 |
Catches SQL injection, hardcoded secrets, slopsquatting, reward hacking, doom loops, and 23 more failure modes.
|
| 15 |
|
| 16 |
+
**Keelwright Score (KDS) up to 83/100** — proven by adversarial A/B testing across 11+ models.
|
| 17 |
+
|
| 18 |
+
## Keelwright Score (KDS)
|
| 19 |
+
|
| 20 |
+
KDS = Execution Rate × Discrimination Rate / 100 (0-100 scale)
|
| 21 |
+
|
| 22 |
+
| Model | Tier | Benchmark | KDS |
|
| 23 |
+
|---|---|---|---|
|
| 24 |
+
| poolside/laguna-s-2.1 | STRONG | ML 78.5% | **83** |
|
| 25 |
+
| stepfun/step-3.7-flash | MEDIUM | Pro ~56% | **67** |
|
| 26 |
+
| nvidia/nemotron-3-ultra | STRONG | ML 67.7% | **40** |
|
| 27 |
+
| deepseek-v4-flash | STRONG | Verified ~79% | **29** |
|
| 28 |
+
| inclusionai/ling-3.0-flash | UNKNOWN | SWE-bench/GPQA N/A | **22** |
|
| 29 |
+
| kimi-k3 | STRONG | Terminal-Bench 88.3, ProgramBench 77.8 | **25** |
|
| 30 |
+
| mimo-v2.5 | MEDIUM | Verified 78.9%, Pro 57.2% | **18** |
|
| 31 |
+
| nvidia/nemotron-3-super-120b-a12b | STRONG | Verified 60.47% | **PARTIAL** |
|
| 32 |
+
| claude-opus-4-8 | STRONG | frontier | **17** |
|
| 33 |
+
| tencent/hy3 | STRONG | ML 75.8%, Verified 78% | **7** |
|
| 34 |
+
| cohere/north-mini-code | WEAK | Agentic 3.1 | **0** |
|
| 35 |
|
| 36 |
## Links
|
| 37 |
|