Finite Interaction Lens: Structured Precision Sharing for Vision-Language Models
Identifying which visual interventions interact in a vision-language model requires comparing their joint and separate effects.
ProofPaper ↗
Key points
- We introduce the Finite Interaction Lens (FIL), a mixed-precision method for evaluating finite intervention families.
- We derive a local cancellation identity and a conditional bound on error propagation through a smooth nonlinearity.
- In one MLP of Qwen3-VL-4B-Instruct, FIL reduces final interaction MSE by 94.4% on 16 synthetic families and by 99.85% and 99.35% in two separate four-image COCO studies, relative to a control that has the same nominal BF16-product budget and higher-precision single interventions but projects each joint candidate independently.
- On these development inputs, FIL improves interaction fidelity at a fixed nominal precision budget.
Sources (1)
- [1]Finite Interaction Lens: Structured Precision Sharing for Vision-Language ModelsarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 18, 10:44 PM
Identifying which visual interventions interact in a vision-language model requires comparing their joint and separate effects.
We introduce the Finite Interaction Lens (FIL), a mixed-precision method for evaluating finite intervention families.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 16, 2026Not Every Mention Is a Claim: Rethinking Object Hallucination Detection in Vision-Language Models
- Oct 16, 2026How Murky Before a VLM Knows It Cannot See? Accuracy, Confidence and Abstention on a Physical Visibility Axis
- Oct 16, 2026Xiaomi-Lens-0 Technical Report
- Oct 16, 2026A Conditional No-Hair Theorem for Transformers: How Language Models Forget Their Prompt
- Oct 16, 2026Towards Looped Models Done Right, Part I: Topology, Input Injection, Recurrent-State Design
- Oct 16, 2026HexaShell: A Thin Harness for Exploring the Limits of VLM Agents as Robot Coding Agents