SMILE: A Smarter, Fairer Metric for Grading Q&A Systems

Kari Jaaskelainen

24 Nov 2025 — 1 min read

How do we fairly score an AI’s answer? Old-school metrics (like ROUGE or Exact Match) reward word overlap, not understanding. LLM judges can “feel” meaning, but they’re costly, inconsistent, and can hallucinate.

Meet SMILE (Semantic Metric Integrating Lexical Exactness): a lightweight way to evaluate answers that blends three signals:

Sentence-level meaning (does the whole answer make sense?)
Keyword-level meaning (are the key ideas there?)
Exact keyword matches (are crucial terms correct?)

This balance captures both what is said and how precisely it’s said—something pure semantics or pure overlap can miss. Across text, image, and video question answering, SMILE aligns strongly with human judgments while staying fast and affordable to run.

Why it matters: better metrics mean more reliable benchmarks, fairer model comparisons, and faster progress—without relying on black-box LLM judges.

Paper: https://arxiv.org/abs/2511.17432v1

Paper: https://arxiv.org/abs/2511.17432v1

Register: https://www.AiFeta.com

#AI #NLP #QuestionAnswering #Evaluation #Metrics #ComputerVision #VQA #LLM #Research

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

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 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

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

Read more

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

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

AI that explains itself—by following the science

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