Both texts below were generated by an LLM with the same prompt. One is watermarked, the other is not. Can you tell them apart?
…
Method #
Each pair was generated by Qwen2.5-32B-Instruct with identical sampling settings. Watermarking is done by SynthID-Text, the method behind both Gemini's and Claude's watermarks. Code, key, prompts, and all raw outputs are in the repo.