LLM Digest

Story

arxiv_llm_reliability · Nov 10, 2026 · paper

Source brief

Sckwoky at SHROOM-Visions: Character-Level Hallucination-Span Detection in Image-Conditioned VLMs via Dual-Prompted LoRA Decoders

arxiv.orgNov 10, 2026
original source linked

In brief

We present our system for the SHROOM-visions 2026 shared task on character-level hallucination span detection in image-conditioned vision-language model (VLM) responses, evaluated across English, French, Italian, and Chinese. The detector is built on a frozen Gemma-4-12B-it backb

Read the original at arxiv.org →Open in live feed Read that day’s brief

Sckwoky at SHROOM-Visions: Character-Level Hallucination-Span Detection in Image-Conditioned VLMs via Dual-Prompted LoRA Decoders

Earlier in this thread 4 items

Post-Training VLMs for Video Mistake Detection

Vendi Anomaly Scores for Efficient and Accurate Anomaly Detection

LogCDM: Interpretable Log Anomaly Detection via Behavioral Graph Drift and Adaptive Signal Fusion

MARS: Multiscale Appearance-Residual Streams for Video Forgery Detection