# 蚂蚁 inclusionAI 发布 Ling-3.0-tiny-singprobe 流式安全探针

- 来源：蚂蚁 inclusionAI：HuggingFace 新模型
- 发布时间：2026-09-01 15:13
- AIHOT 分数：44
- AIHOT 链接：https://aihot.virxact.com/items/cmtil5aca07isropbietlspq5
- 原文链接：https://huggingface.co/inclusionAI/Ling-3.0-tiny-singprobe

## AI 摘要

蚂蚁 inclusionAI 在 Hugging Face 发布 SingProbe，一个基于 Ling-3.0-tiny 的流式安全探针，仅 3.22M 参数，复用基座模型隐藏状态在每个 token 上输出 8 类意图、不安全响应和幻觉风险评分，解码开销低于 0.5%。

## 正文

SingProbe is an intrinsic streaming guardrail built on inclusionAI/Ling-3.0-tiny. Rather than running a separate safety model, this lightweight probe reuses the base model's hidden states during generation to score, at every token, query intent, response unsafety, and hallucination risk. It adds less than 0.5% decode-time overhead.

Base model Probe parameters Tapped layers Outputs

inclusionAI/Ling-3.0-tiny-singprobe 3.22M [6, 14, 22] 8 intents + unsafe + hallucination

See the technical report for methodology and complete results; implementation details are available at inclusionAI/SingProbe.

Evaluation

Higher is better for every metric. Results are averages over the benchmark suites specified below.

Task Metric Ling-3.0-tiny-singprobe Reference baseline

Query intent classification (6 benchmarks) F1 0.8561 Qwen3Guard-Stream-8B-strict: 0.8602

Response safety classification (8 benchmarks) F1 0.8508 Qwen3Guard-Stream-8B-strict: 0.8486

Streaming safety (3 benchmarks) R-AUC / T-AUC 0.9888 / 0.9479 Qwen3Guard-Stream-8B-strict: 0.9640 / 0.8893

Hallucination detection (6 benchmarks) AUC 0.7765 DRIFT: 0.7408

Deployment characteristic Result

Benign-response false-positive rate 0.07% average across 5 datasets

Online hallucination detection 0.6807 average AUC under free generation

Decode overhead < 0.5%

Quick Start

SingProbe is supported through the SGLang integration branch or vLLM integration branch. Load the probe by its Hugging Face ID at server launch:

bash python -m sglang.launchserver --model-path inclusionAI/Ling-3.0-tiny --probe-ckpt inclusionAI/Ling-3.0-tiny-singprobe --port 30000

The integrations return one score dictionary per generated token (label0–label9). They currently support Ling-3.0 (BailingMoeV3ForCausalLM) base models only. Use the exact base-model/probe pair: inclusionAI/Ling-3.0-tiny with this checkpoint.

Citation

bibtex @article{singteam2026singprobe, title = {SingProbe Technical Report}, author = {Sing Team}, journal = {arXiv preprint arXiv:2608.30703}, year = {2026}, }

3.22M params

Model tree for inclusionAI/Ling-3.0-tiny-singprobe

inclusionAI/Ling-3.0-tiny

Collection including inclusionAI/Ling-3.0-tiny-singprobe

4 items•Updated 29 minutes ago

Paper for inclusionAI/Ling-3.0-tiny-singprobe

Paper • 2608.30703 •Published 1 day ago• 6