smollm2-f5bd7c44·2 events·first seen Aliases: SmolLM2
Researchers introduce PRISM, a single-layer logistic probe over LLM hidden states that distinguishes physically grounded danger (e.g., unsafe robot actions) from ordinary text-level content danger, showing these are separable signals in representations across multiple model families. PRISM achieves 86.2–87.7% accuracy on SafeAgentBench with substantially lower false-positive rates than same-scale LLM judges, which over-block safe tasks at 24.7–39.0% FPR. The authors also release PhysicalSafetyBench-1K (PSB-1K), a 1,000-pair contrastive benchmark for evaluating physical-risk detection without relying on explicit harm keywords. The work is relevant to safety in embodied AI and agentic systems where linguistic safety filters are insufficient.
Hugging Face published a blog post introducing Smol2Operator, a post-training approach for building GUI agents capable of computer use tasks. The work focuses on training small language models to operate graphical user interfaces, extending the SmolLM2 model family into the agent/computer-use domain. The post likely covers training methodology, datasets, and evaluation of the resulting GUI agent capabilities.