AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Latent Speech-Text Transformer

Yen-Ju Lu, Yashesh Gaur, Wei Zhou, Benjamin Muller, Jesus Villalba, Najim Dehak, Luke Zettlemoyer, Gargi Ghosh, Mike Lewis, Srini Iyer, Duc Le

ICLR 2026 · Image generation › Autoregressive and latent generation · approach: Neural text-to-speech synthesis · 5 citations (OpenAlex)

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
add speech patching (curriculum) to the speech-text LMBase SpeechLLM · HellaSwag S->S (compute-controlled)accuracy: 39.0 -> 45.5Improved
add speech patching (curriculum) to the speech-text LMBase SpeechLLM · HellaSwag T->T (compute-controlled)accuracy: 47.0 -> 52.2Improved
add speech patching (curriculum) with data fixedBase SpeechLLM · HellaSwag (data-controlled)accuracy: S->S 40.2 -> 45.5; T->T 49.6 -> 52.2Improved
replace patching with BPE on speech tokensBase SpeechLLM · HellaSwag S->S (compute-controlled)accuracy: 39.0 -> 38.0Hurt
use curriculum rather than aligned-only patching (ablation)LST · HellaSwag S->S (compute-controlled)accuracy: 42.7 -> 45.5Improved
use aligned patching instead of static at matched size 6 (ablation)LST · StoryCloze S->Saccuracy: 58.7 -> 60.3Improved
use aligned (sil merged) instead of static at matched size 9 (ablation)LST · HellaSwag S->Saccuracy: 37.2 -> 38.5Improved
add curriculum patching vs static size 4 (ablation)LST · HellaSwag S->Saccuracy: 40.5 -> 41.3Improved
scale to 1.8B with LST (compute-optimal)Baseline · HellaSwag S->Saccuracy: 35.3 -> 39.0Improved
scale to 420M with LST (compute-optimal)Baseline · HellaSwag S->Saccuracy: 28.4 -> 29.2Improved
add LST at 7B fixed tokensBaseline · HellaSwag S->Saccuracy: 42.0 -> 44.2Improved
add LST for ASR adaptationBaseline · LibriSpeech clean ASR, 1k itersWER %: 140 -> 6.8Improved

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