SingProbe Technical Report
We present SingProbe, an open intrinsic guardrail framework for generation-time monitoring of LLMs. Intrinsic guardrails reuse hidden states already produced by the base model during autoregressive decoding, rather than relying on an independent model to repeatedly process generated text. While this route has been explored in industrial systems, the community lacks a broadly reusable open stack that combines cross-model guard adaptations, unified training methods, serving integrations, and systematic evaluation resources. SingProbe is designed to provide this missing layer and uses a lightweight probe to continuously produce query-intent, response-safety, and hallucination-risk signals du...
| Authors | Sing Team |
|---|---|
| Primary category | cs.CR |
| Published | 31 August 2026 |
| Last updated at source | 21 September 2026 |
Models and datasets referencing this paper
- inclusionAI/Qwen3-0.6B-singprobeModel
- inclusionAI/Qwen3.6-35B-A3B-singprobeModel
- inclusionAI/Qwen3.5-2B-singprobeModel
- inclusionAI/Ling-3.0-tiny-singprobeModel
- inclusionAI/Llama-3.2-1B-Instruct-singprobeModel
- inclusionAI/Ling-3.0-flash-singprobeModel
- inclusionAI/gpt-oss-20b-singprobeModel
- inclusionAI/Qwen3.8-27B-singprobeModel
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.