SmolLM
PulseAugur coverage of SmolLM — every cluster mentioning SmolLM across labs, papers, and developer communities, ranked by signal.
- 2026-09-03 product_launch A new speech model based on the SmoLLM backbone was released, featuring 80ms latency for natural conversations. source
2 day(s) with sentiment data
-
Prompt echoing in small LLMs linked to induction heads, not just data leakage
Researchers have investigated the phenomenon of prompt echoing in small instruction-following language models. They analyzed models from various families, including Gemma, Llama, Qwen, SmolLM, and OLMo, to understand wh…
-
New SmoLLM speech model achieves 80ms latency for natural conversations
A new speech model, built on a 135 million parameter SmoLLM backbone, has been developed to facilitate natural, full-duplex conversations. This model achieves an impressive 80ms latency, allowing for smooth replies and …
-
New IAR framework enhances LLM document knowledge internalization
Researchers have developed a new three-stage post-training framework called IAR (Inject, Align, Recover) designed to improve how large language models internalize knowledge from specific documents for retrieval-free que…
-
New method estimates LLM training data composition from vocabularies
Researchers have developed a new method called Quantile-Guided Density Estimation (QGDE) to estimate the composition of hidden training corpora for large language models (LLMs). This technique leverages released tokeniz…