AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Gogo: Group-wise granularity-ordered codec for stable and efficient speech generation

Weidong Chen, Helen M. Meng, Xixin Wu

ICLR 2026 · Image generation › Autoregressive and latent generation · approach: Neural text-to-speech synthesis

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use group-wise quantization instead of frame-wise RVQ (token position 1)Gogo tokens · AR perplexityperplexity: 2.3 -> 0.9Improved
use group-wise quantization instead of frame-wise RVQ (token position 10)Gogo tokens · AR perplexityperplexity: 691.4 -> 228.3Improved
use group-wise instead of single-layer frame-wise VQ for the coarsest tokenGogo tokens · AR perplexityperplexity: 247.8 -> 0.9Improved
add GRPO token allocator to GogoSpeech (47 -> 36 tokens/s) (ablation)GogoSpeech · Seed-TTS test-enRTF: 0.535 -> 0.455Improved
add GRPO token allocator to GogoSpeech (ablation)GogoSpeech · Seed-TTS test-enWER: 2.394 -> 2.469Hurt
add GRPO token allocator to GogoSpeech (ablation)GogoSpeech · Seed-TTS test-enSMOS: 4.381 -> 4.253Hurt
add GRPO token allocator to GogoSpeech (ablation)GogoSpeech · Seed-TTS test-en long-formWER: 1.788 -> 1.845Hurt
use two-stage coarse-to-fine generation (GogoSpeech) versus F5-TTS on long-form speechGogoSpeech · Seed-TTS test-en long-formWER: 1.812 -> 1.788Improved
use Gogo at 47 TPS vs X-codec2 at 50 TPSGogo · LibriTTS test-cleanWER: 6.40 -> 6.35Improved
retain more than six tokens per group at reconstructionGogo · LibriTTS test-cleanWERNo clear effect
retain more than four tokens per group at reconstructionGogo · LibriTTS test-cleanPESQImproved
move probing from coarse token 1 to fine token 10 for pitchGogo tokens · LibriTTS probingprobe MSE (pitch): 1455 -> 1012Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy