Scene character-rewrite
marvel-universe-villains-sun-tzu · Dialogue Punch-Up
Scores by Model
| Model | Score | Judge | When |
|---|---|---|---|
| qwen/qwen3-8b | 0.870 | openai/gpt-5-mini | Sept. 24, 2025, 8:44 p.m. |
| [email protected]/Qwen3-8B-b0d7af1f | 0.720 | meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo | Oct. 29, 2025, 9:19 a.m. |
| google/gemma-3-12b-it | 0.631 | openai/gpt-5-mini | Sept. 24, 2025, 9:10 p.m. |
| microsoft/phi-3.5-mini-128k-instruct | 0.572 | openai/gpt-5-mini | Sept. 24, 2025, 9:13 p.m. |
| deepseek/deepseek-r1-distill-qwen-14b | 0.548 | openai/gpt-5-mini | Sept. 24, 2025, 9:16 p.m. |
| [email protected]/Qwen3-8B-da5790fa | 0.443 | meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo | Oct. 10, 2025, 11:29 a.m. |
| mistralai/mistral-7b-instruct | 0.367 | openai/gpt-5-mini | Sept. 24, 2025, 8:45 p.m. |
| google/gemini-2.5-flash | 0.355 | openai/gpt-5-mini | Sept. 24, 2025, 8:35 p.m. |
| qwen/qwen-2.5-7b-instruct | 0.335 | openai/gpt-5-mini | Sept. 24, 2025, 8:37 p.m. |
| [email protected]/Qwen2.5-7B-Instruct-521d3af9 | 0.308 | meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo | Oct. 10, 2025, 8:49 a.m. |
| [email protected]/Qwen3-14B-e66d90ff | 0.210 | meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo | Oct. 10, 2025, 1:35 p.m. |
| qwen/qwen3-14b | 0.054 | openai/gpt-5-mini | Sept. 24, 2025, 8:40 p.m. |
| microsoft/phi-3-medium-128k-instruct | 0.043 | openai/gpt-5-mini | Sept. 24, 2025, 9:13 p.m. |
| meta-llama/llama-3.1-8b-instruct | 0.000 | openai/gpt-5-mini | Sept. 24, 2025, 9:09 p.m. |
| neversleep/noromaid-20b | 0.000 | openai/gpt-5-mini | Sept. 24, 2025, 9:20 p.m. |
| [email protected]/Qwen3-14B-984c85c4 | 0.000 | meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo | Oct. 31, 2025, 8:54 a.m. |