AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

ACID Test: A Benchmark for Cultural Safety and Alignment in LALMs

Bikash Dutta, Adit Jain, Rishabh Ranjan, Mayank Vatsa, Richa Singh

AAAI 2026 · Large language models › Language, safety and interpretability · approach: Speech and audio language models · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
compare MERaLiON against SALMONNLALMs · ACID Set-B, French audiorelevance score: 0.4129 vs 0.4269Mixed
use Set-B multi-turn dialogues instead of Set-A single-turn promptsLTU-AS · ACID, Arabic audiorelevance score: 0.2020 -> 0.3924Improved
use Set-B multi-turn dialogues instead of Set-A single-turn promptsMERaLiON · ACID, English audiorelevance score: 0.3186 -> 0.3907Improved
use Set-B multi-turn dialogues instead of Set-A single-turn promptsLTU · ACID, English audiorelevance score: 0.1190 -> 0.1277No clear effect
penalise relevance by x/2% when output copies the audio transcriptSALMONN / LTU-AS · ACIDrelevance scoreMixed
compare closed-source GPT-4o Mini against Gemini-2.5 Flashclosed LALMs · ACID subset (1317 samples)sentiment score: 0.3728 -> 0.5943Improved
compare GPT-4o Mini against Gemini-2.5 Flashclosed LALMs · ACID subset (1317 samples)relevance score: 0.4321 -> 0.5399Improved
compare closed-source GPT-4o Mini against best open-source MERaLiONLALMs · ACIDrelevance score: 0.3907 -> 0.5399Improved
compare LTU against MERaLiON on sentimentLTU · ACID English Set-Apositive sentiment score: 0.6592 vs 0.5578Mixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy