Lost in Translation and Noise: Why VLMs Stumble on Real-World Tables

Kari Jaaskelainen

24 Nov 2025 — 1 min read

Does your vision-language model crush table QA benchmarks—until the tables look like scans or switch languages? MirageTVQA puts that to the test.

It’s a new benchmark of ~60,000 question–answer pairs across 24 languages, with tables that are multilingual and visually imperfect (think scanned documents with noise), reflecting how data appears in the wild.

Noise hurts—badly: leading VLMs lose over 35% performance when tables include realistic visual artifacts.
English-first bias: reasoning that works in English often fails to transfer to other languages.

Why this matters: real invoices, reports, and forms are messy and multilingual. MirageTVQA is a yardstick to build and measure more robust table reasoning.

Paper: https://arxiv.org/abs/2511.17238v1 • Data & code: https://github.com/anshulsc/MirageTVQA

Paper: https://arxiv.org/abs/2511.17238v1

Register: https://www.AiFeta.com

#AI #VLM #VisionLanguage #Multilingual #OCR #TableQA #Benchmark #MLResearch #RobustAI #DocumentAI

Read more

Meet GRACE: a moral governor for safer, more transparent AI

Meet GRACE: a moral governor for safer, more transparent AI

AI agents are getting powerful—so how do we make sure they do the right thing, not just the effective thing? Meet GRACE, a reason-based moral governor that keeps AI behavior aligned with human norms by separating moral reasoning from goal-driven decision-making. * Moral Module: uses deontic logic and explicit reasons

Why some fine-tuned LLMs miss phishing—and how to fix it

Why some fine-tuned LLMs miss phishing—and how to fix it

Not all fine-tuned LLMs spot phishing equally. A new study tests Llama 3.1 8B, Gemma 2 9B, and Mistral on high-stakes phishing detection—and uses SHAP and mechanistic interpretability to reveal why models do (or don’t) generalize. * Architecture × data diversity matters: Gemma 2 9B hits state-of-the-art performance (F1

AI that explains itself—by following the science

AI that explains itself—by following the science

AI that explains itself—by following the science Black-box AI can be powerful, but it often can’t tell us why it made a decision. Concept Bottleneck Models (CBMs) try to fix that by predicting human-understandable concepts first, then the final answer. The catch: standard CBMs ignore domain-specific cause-and-effect and

Meet GenomAgent: A Team of AI Specialists for Smarter Genomics Q&A

Meet GenomAgent: A Team of AI Specialists for Smarter Genomics Q&A

TL;DR: Finding reliable answers in genomics is hard. GenomAgent turns one big AI into a coordinating team of specialists—and beats the current leader by 12% on a key benchmark. Genomic facts live across many databases. Standard chatbots struggle because they can’t flexibly query those sources. GeneGPT added