3 ms·
tldr - Qwen 3.5 9B finetuned on a "simple secret dataset" (500 million tokens of Claude 4.8 Opus, Claude 5 Fable, GPT 5.5 and similar models)
by zihotki 3mo ago
tldr - Qwen 3.5 9B finetuned on a "simple secret dataset" (500 million tokens of Claude 4.8 Opus, Claude 5 Fable, GPT 5.5 and similar models)