Search papers, labs, and topics across Lattice.
Affiliation:
5
1
9
6
A deployable system for streaming Thai ASR that lets a user steer its vocabulary at decode time, without retraining is presented, which lifts keyword recall from 16.6% to 20.7% at no accuracy cost and negligible overhead.
It is found that synthetic-only training can be competitive, and the 0.9B-parameter PaddleOCR-VL-1.6 is adapted into Wayu-Paxa-OCR-Zero, a Thai OCR model adapted without OCR labels from real Thai document pages, showing that synthetic-only training can be competitive.
The resulting 82M-parameter model, Wayu-Paxa-TTS-Edge, enables on-device Thai TTS without reference audio and achieves the lowest pause-placement error and intra-word pause rates among the three systems.
Current LLM safety measures are critically vulnerable to attacks grounded in Thai cultural nuances, as demonstrated by a new benchmark showing higher attack success rates compared to general Thai-language attacks.
Forget fancy prompt optimization: carefully crafted rule-based role prompts with enforced function calls dramatically improve the performance of tool-augmented LLM dialogue agents.
