Paper
Auditing Reciprocal Sentiment Alignment: Inversion Risk, Dialect Representation and Intent Misalignment in Transformers
Authors
Nusrat Jahan Lia, Shubhashis Roy Dipta
Abstract
The core theme of bidirectional alignment is ensuring that AI systems accurately understand human intent and that humans can trust AI behavior. However, this loop fractures significantly across language barriers. Our research addresses Cross-Lingual Sentiment Misalignment between Bengali and English by benchmarking four transformer architectures. We reveal severe safety and representational failures in current alignment paradigms. We demonstrate that compressed model (mDistilBERT) exhibits 28.7% "Sentiment Inversion Rate," fundamentally misinterpreting positive user intent as negative (or vice versa). Furthermore, we identify systemic nuances affecting human-AI trust, including "Asymmetric Empathy" where some models systematically dampen and others amplify the affective weight of Bengali text relative to its English counterpart. Finally, we reveal a "Modern Bias" in the regional model (IndicBERT), which shows a 57% increase in alignment error when processing formal (Sadhu) Bengali. We argue that equitable human-AI co-evolution requires pluralistic, culturally grounded alignment that respects language and dialectal diversity over universal compression, which fails to preserve the emotional fidelity required for reciprocal human-AI trust. We recommend that alignment benchmarks incorporate "Affective Stability" metrics that explicitly penalize polarity inversions in low-resource and dialectal contexts.
Metadata
Related papers
Vibe Coding XR: Accelerating AI + XR Prototyping with XR Blocks and Gemini
Ruofei Du, Benjamin Hersh, David Li, Nels Numan, Xun Qian, Yanhe Chen, Zhongy... • 2026-03-25
Comparing Developer and LLM Biases in Code Evaluation
Aditya Mittal, Ryan Shar, Zichu Wu, Shyam Agarwal, Tongshuang Wu, Chris Donah... • 2026-03-25
The Stochastic Gap: A Markovian Framework for Pre-Deployment Reliability and Oversight-Cost Auditing in Agentic Artificial Intelligence
Biplab Pal, Santanu Bhattacharya • 2026-03-25
Retrieval Improvements Do Not Guarantee Better Answers: A Study of RAG for AI Policy QA
Saahil Mathur, Ryan David Rittner, Vedant Ajit Thakur, Daniel Stuart Schiff, ... • 2026-03-25
MARCH: Multi-Agent Reinforced Self-Check for LLM Hallucination
Zhuo Li, Yupeng Zhang, Pengyu Cheng, Jiajun Song, Mengyu Zhou, Hao Li, Shujie... • 2026-03-25
Raw Data (Debug)
{
"raw_xml": "<entry>\n <id>http://arxiv.org/abs/2602.17469v1</id>\n <title>Auditing Reciprocal Sentiment Alignment: Inversion Risk, Dialect Representation and Intent Misalignment in Transformers</title>\n <updated>2026-02-19T15:35:13Z</updated>\n <link href='https://arxiv.org/abs/2602.17469v1' rel='alternate' type='text/html'/>\n <link href='https://arxiv.org/pdf/2602.17469v1' rel='related' title='pdf' type='application/pdf'/>\n <summary>The core theme of bidirectional alignment is ensuring that AI systems accurately understand human intent and that humans can trust AI behavior. However, this loop fractures significantly across language barriers. Our research addresses Cross-Lingual Sentiment Misalignment between Bengali and English by benchmarking four transformer architectures. We reveal severe safety and representational failures in current alignment paradigms. We demonstrate that compressed model (mDistilBERT) exhibits 28.7% \"Sentiment Inversion Rate,\" fundamentally misinterpreting positive user intent as negative (or vice versa). Furthermore, we identify systemic nuances affecting human-AI trust, including \"Asymmetric Empathy\" where some models systematically dampen and others amplify the affective weight of Bengali text relative to its English counterpart. Finally, we reveal a \"Modern Bias\" in the regional model (IndicBERT), which shows a 57% increase in alignment error when processing formal (Sadhu) Bengali. We argue that equitable human-AI co-evolution requires pluralistic, culturally grounded alignment that respects language and dialectal diversity over universal compression, which fails to preserve the emotional fidelity required for reciprocal human-AI trust. We recommend that alignment benchmarks incorporate \"Affective Stability\" metrics that explicitly penalize polarity inversions in low-resource and dialectal contexts.</summary>\n <category scheme='http://arxiv.org/schemas/atom' term='cs.CL'/>\n <category scheme='http://arxiv.org/schemas/atom' term='cs.HC'/>\n <published>2026-02-19T15:35:13Z</published>\n <arxiv:primary_category term='cs.CL'/>\n <author>\n <name>Nusrat Jahan Lia</name>\n </author>\n <author>\n <name>Shubhashis Roy Dipta</name>\n </author>\n </entry>"
}