# nanoparticles

Published articles for nanoparticles.

This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles.

## Do Multimodal Models Decide Consistently on Equivalent Optical Spectra?

DevFeed: [Do Multimodal Models Decide Consistently on Equivalent Optical Spectra?](<https://devfeed.tech/articles/do-multimodal-models-decide-consistently-on-equivalent-optical-spectra-114382.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.00428>)

Author: Ziyuan Wang, Genliang Zhu, Xiaobo Chen

Published: 2026-11-03T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Large Language Model](<https://devfeed.tech/topics/llm.md>), [Fine-tuning](<https://devfeed.tech/topics/fine-tuning.md>), [lora](<https://devfeed.tech/topics/lora.md>)

Tags: [cs-cl](<https://devfeed.tech/tags/cs-cl.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [fine-tuning](<https://devfeed.tech/tags/fine-tuning.md>), [gpt-6-luna](<https://devfeed.tech/tags/gpt-6-luna.md>), [invariance](<https://devfeed.tech/tags/invariance.md>), [lora](<https://devfeed.tech/tags/lora.md>), [metasurfaces](<https://devfeed.tech/tags/metasurfaces.md>), [multimodal](<https://devfeed.tech/tags/multimodal.md>), [multimodal-models](<https://devfeed.tech/tags/multimodal-models.md>), [nanoparticles](<https://devfeed.tech/tags/nanoparticles.md>), [optical](<https://devfeed.tech/tags/optical.md>), [physics-optics](<https://devfeed.tech/tags/physics-optics.md>), [qwen3-5](<https://devfeed.tech/tags/qwen3-5.md>), [silicon](<https://devfeed.tech/tags/silicon.md>), [spectrum](<https://devfeed.tech/tags/spectrum.md>)

### AI overview

Across measured optical spectra of silicon metasurfaces and gold nanoparticles, equivalent representations can produce sharply different model decisions: changing format shifts accuracy by up to 41 percentage points, while asking in photon energy helps some model-view pairs and hurts others. Unadapted models rarely abstain when the relevant spectral extremum is missing. LoRA fine-tuning improves d3-nano on trained table formats, but can teach it to abstain whenever input is truncated. Including answerable truncations avoids that shortcut across six seeds and produces a 90-100 point gap in abstention between cuts that remove queried evidence and cuts that do not; extra original-task data generally does not. The two training sets perform similarly on gold spectra. The abstract argues for testing representation consistency and missing-evidence abstention alongside average accuracy.

### Source excerpt

arXiv:2611.00428v1 Announce Type: new Abstract: A spectrum can be given to a model as a table in nanometers or as a plot in photon energy, and the decision should not depend on which one it sees. When the data needed for an answer are missing, the model should say so. We test representation invariance and abstention on measured optical spectra of silicon metasurfaces (MOCLIP) and gold nanoparticles, with labels computed from the data. Each spectrum is shown as four tables (nm, $\mu$m, scientific notation, eV) and two plots (nm, eV), and partial spectra remove data inside or outside the queried range. We evaluate the decision models d3-nano and Jev 1.13, three GPT-6 Luna configurations and the Qwen3.5-2B base model. Changing only the representation moves accuracy by up to 41 percentage points within one model, mostly on photon-energy tables and plots. With the same evidence, asking in photon energy instead of wavelength helps some model-view pairs by 16-50 points and hurts others by up to 25. When the relevant extremum is removed, unadapted models rarely answer "undetermined". LoRA fine-tuning of d3-nano improves comparisons on the trained table formats, but much of its gain on unanswerable questions comes from a shortcut: it abstains whenever the input is truncated. Training with answerable truncations avoids the shortcut in all six seeds, with a 90-100 point gap between abstention after inside and outside cuts on held-out structures; with the same amount of extra original-task data, five of six seeds keep the shortcut. On gold spectra the two training sets perform similarly. Multimodal language models used for scientific decisions should be tested on equivalent representations and on missing evidence, not only on average accuracy. We release the tasks, per-response results and analysis code.