Anika Sharma

Alice Zhong, Phoebe Chen, Anika Sharma et al.

Sexual and reproductive health (SRH) remains shaped by structural barriers that leave many without judgment-free information. AI chatbots offer anonymous alternatives, but access alone does not ensure equity when socioeconomic determinants shape whose capabilities these tools expand or constrain. Conventional methods for evaluating human-AI interaction were not designed to capture whether technologies holistically support reproductive autonomy. We introduce CARE, Capability Approach for Reproductive Equity, developing capabilities, functionings, and conversion factors into a Normative Design Lens and an Evaluation Lens for AI in SRH contexts. Evaluating SRH-specific non-LLM chatbots, general-use LLMs, and search engine features along credibility and reasoning, we identify two epistemic harms: source opacity and response rigidity. We conclude with design and evaluation recommendations, participatory auditing strategies, and policy implications for high-stakes domains where AI intersects with inequity.

Anika Sharma, Tianyi Niu, Emma Wrenn et al.

The phenomenon of sound symbolism, the non-arbitrary mapping between word sounds and meanings, has long been demonstrated through anecdotal experiments like Bouba Kiki, but rarely tested at scale. We present the first computational cross-linguistic analysis of sound symbolism in the semantic domain of size. We compile a typologically broad dataset of 810 adjectives (27 languages, 30 words each), each phonemically transcribed and validated with native-speaker audio. Using interpretable classifiers over bag-of-segment features, we find that phonological form predicts size semantics above chance even across unrelated languages, with both vowels and consonants contributing. To probe universality beyond genealogy, we train an adversarial scrubber that suppresses language identity while preserving size signal (also at family granularity). Language prediction averaged across languages and settings falls below chance while size prediction remains significantly above chance, indicating cross-family sound-symbolic bias. We release data, code, and diagnostic tools for future large-scale studies of iconicity.

Anika Sharma, Malavika Mampally, Chidaksh Ravuru et al.

As Large Language Models (LLMs) increasingly mediate stigmatized health decisions, their capacity to understand complex psychological phenomena remains inadequately assessed. Can LLMs understand what we cannot say? We investigate whether LLMs coherently represent abortion stigma across cognitive, interpersonal, and structural levels. We systematically tested 627 demographically diverse personas across five leading LLMs using the validated Individual Level Abortion Stigma Scale (ILAS), examining representation at cognitive (self-judgment), interpersonal (worries about judgment and isolation), and structural (community condemnation and disclosure patterns) levels. Models fail tests of genuine understanding across all dimensions. They underestimate cognitive stigma while overestimating interpersonal stigma, introduce demographic biases assigning higher stigma to younger, less educated, and non-White personas, and treat secrecy as universal despite 36% of humans reporting openness. Most critically, models produce internal contradictions: they overestimate isolation yet predict isolated individuals are less secretive, revealing incoherent representations. These patterns show current alignment approaches ensure appropriate language but not coherent understanding across levels. This work provides empirical evidence that LLMs lack coherent understanding of psychological constructs operating across multiple dimensions. AI safety in high-stakes contexts demands new approaches to design (multilevel coherence), evaluation (continuous auditing), governance and regulation (mandatory audits, accountability, deployment restrictions), and AI literacy in domains where understanding what people cannot say determines whether support helps or harms.