Meta paper proposes TextSeal2: merging dual keys improves AI text watermark detection rate
Meta's new paper proposes TextSeal2, which fuses two keys into one virtual key and adds new noise, so that every token carries the complete watermark signal while maintaining answer diversity. For speculative decoding, it selects replacement tokens from the remaining candidates of the draft's random competition, keeping the acceptance rate unchanged. On Qwen3.5-9B with speculative decoding enabled and diversity held equal, TextSeal2 achieves a detection rate of 89% on 200-token responses at a 0.1% false positive rate, compared with 71% for TextSeal and 67% for textGrain.
Why read this
AI score 55
Meta's new paper merges dual keys into a single virtual key, maintaining the detection signal under speculative decoding and providing a same-condition comparison with TextSeal and textGrain.
Topics
Body
New Meta paper shows that blending 2 secret keys into 1 makes hidden watermarks in AI-written text easier to detect, even when the AI uses a common trick (like speculative decoding) to write faster.
Watermarks hide a secret-key signal in which tokens an LLM samples, so a detector can later flag its text. Meta's earlier TextSeal picked 1 of 2 keys per token at random, so detection had to average both and lost signal.
TextSeal2 blends both keys into 1 virtual key plus fresh noise, so answers stay varied and every token carries full signal. For speculative decoding, it picks replacement tokens from the leftovers of the drafts' random races, so acceptance rates stay unchanged.
On Qwen3.5-9B with speculative decoding and equal diversity, it detected 89% of 200-token answers at a 0.1% false-positive rate. TextSeal caught 71% and textGrain 67%, at the same acceptance rate.