Researchers have developed Gacha Decoding, a new method for generating diverse language model outputs that scales with model capability, outperforming existing approaches in various domains.
We introduce Gacha Decoding, an inference-time method for eliciting diverse language model generations that scales with model capability.
Abstract: We introduce Gacha Decoding, an inference-time method for eliciting diverse language model generations that scales with model capability. Across open-ended domains (in-the-wild chat, creative writing, planning for image generation, and protein design), Gacha Decoding significantly outperforms existing generation diversity approaches at equal quality (up to 2.4x Vendi over the next-best…
