Speech-language models
PulseAugur coverage of Speech-language models — every cluster mentioning Speech-language models across labs, papers, and developer communities, ranked by signal.
-
Speech Language Models Fail to Grasp Sound Symbolism Like Humans
A new research paper explores whether speech language models (SLMs) can understand sound symbolism, the human ability to associate speech sounds with perceptual qualities like sharpness or roundness. The study found tha…
-
New vLLM pipeline unifies audio generation and understanding
Researchers have developed a novel inference pipeline utilizing vLLM to unify audio understanding and generation tasks. This system addresses the challenges of high-throughput multimodal generation, particularly for spe…
-
New SpeechEQ benchmark evaluates AI emotional intelligence in voice models
Researchers have introduced SpeechEQ, a new framework designed to evaluate the emotional intelligence of speech-language models (SLMs). This framework includes a dataset of 2,265 dialogues and a multi-turn evaluation pr…
-
Speech-language models implicitly transcribe spoken words, study finds
A new research paper published on arXiv explores the internal workings of interleaved speech-language models (SLMs). The study reveals that these models, even when not explicitly trained for speech recognition, undergo …
-
Speech language models vulnerable to backdoor attacks
Researchers have analyzed how backdoor attacks propagate through speech language models, which are complex systems composed of multiple interconnected components. Their findings indicate that backdoors can spread throug…
-
ParaBridge method improves speech models' paralinguistic understanding
Researchers have developed ParaBridge, a novel on-policy self-distillation method designed to improve speech language models' ability to incorporate paralinguistic cues into dialogue. This technique trains models to bet…