Agam Goyal

Drishti Goel, Agam Goyal, Veda Duddu et al.

Language models are increasingly being deployed for conversational support in informal caregiving contexts, where interactions often extend beyond information-seeking: caregivers seek emotional reassurance, guidance, and help, while navigating uncertain, relationally complex care decisions. Yet most safety evaluations assess model behavior under generic prompts, leaving a critical question unexamined: does a model's safety profile change with its support role? We study this by operationalizing four expert-reviewed support roles grounded in social support theory: Inform, Coach, Relate, and Listen, and comparing them against two baseline controls: a basic prompting condition and a retrieval-augmented generation (RAG) condition. We evaluate across three language models (GPT-4o-mini, Llama-3.1-8B-Instruct, and MedGemma-1.5-4b-it) on 5,000 real-world queries from online Alzheimer's Disease and Related Dementias (ADRD) communities. We find that the LLM's support role systematically shapes both the prevalence and composition of interactional risks. Furthermore, a human evaluation study reveals a perceived quality--safety tension: more directive, information-oriented roles are rated as more helpful and trustworthy despite exhibiting elevated interactional risk profiles. We release ~90,000 support role-conditioned model responses with risk annotations as an ecologically grounded resource for research on safer LLM-mediated conversational support.

Soorya Ram Shimgekar, Agam Goyal, Amruta Parulekar et al.

Large language models (LLMs) are increasingly deployed in conversational settings where user tone ranges from polite to adversarial or toxic, yet less is known about whether toxic language in otherwise semantically equivalent prompts can degrade factual reliability. We study how lexical and tone-based prompt perturbations affect the factual reliability of LLMs. Using controlled prompt variations across polite, random, and three toxicity levels, we evaluate five LLMs on ARC-Easy, GSM8K, and MMLU. We find that toxic lexical perturbations consistently reduce factual accuracy and increase uncertainty, while polite phrasing yields limited and inconsistent changes. To examine whether these answer inconsistencies correspond to internal changes, we conduct attribution-graph analyses of model activations and influences. We find that increasing toxicity selectively amplifies perturbation-sensitive variant nodes while relatively stable core reasoning nodes remain more invariant. These findings position prompt tone as a critical dimension of LLM reliability and provide behavioral and mechanistic evidence that surface-level lexical variation can alter factual outputs and internal computation.

Yun-Shiuan Chuang, Agam Goyal, Nikunj Harlalka et al.

Accurately simulating human opinion dynamics is crucial for understanding a variety of societal phenomena, including polarization and the spread of misinformation. However, the agent-based models (ABMs) commonly used for such simulations often over-simplify human behavior. We propose a new approach to simulating opinion dynamics based on populations of Large Language Models (LLMs). Our findings reveal a strong inherent bias in LLM agents towards producing accurate information, leading simulated agents to consensus in line with scientific reality. This bias limits their utility for understanding resistance to consensus views on issues like climate change. After inducing confirmation bias through prompt engineering, however, we observed opinion fragmentation in line with existing agent-based modeling and opinion dynamics research. These insights highlight the promise and limitations of LLM agents in this domain and suggest a path forward: refining LLMs with real-world discourse to better simulate the evolution of human beliefs.

Agam Goyal, Charlotte Lambert, Yoshee Jain et al.

A major task for moderators of online spaces is norm-setting, essentially creating shared norms for user behavior in their communities. Platform design principles emphasize the importance of highlighting norm-adhering examples and explicitly stating community norms. However, norms and values vary between communities and go beyond content-level attributes, making it challenging for platforms and researchers to provide automated ways to identify desirable behavior to be highlighted. Current automated approaches to detect desirability are limited to measures of prosocial behavior, but we do not know whether these measures fully capture the spectrum of what communities value. In this paper, we use upvotes, which express community approval, as a proxy for desirability and examine 16,000 highly-upvoted comments across 80 popular sub-communities on Reddit. Using a large language model, we extract values from these comments across two years (2016 and 2022) and compile 64 and 72 $\textit{macro}$, $\textit{meso}$, and $\textit{micro}$ values for 2016 and 2022 respectively, based on their frequency across communities. Furthermore, we find that existing computational models for measuring prosociality were inadequate to capture on average $82\%$ of the values we extracted. Finally, we show that our approach can not only extract most of the qualitatively-identified values from prior taxonomies, but also uncover new values that are actually encouraged in practice. Our findings highlight the need for nuanced models of desirability that go beyond preexisting prosocial measures. This work has implications for improving moderator understanding of their community values and provides a framework that can supplement qualitative approaches with larger-scale content analyses.

Yun-Shiuan Chuang, Siddharth Suresh, Nikunj Harlalka et al.

Human groups are able to converge on more accurate beliefs through deliberation, even in the presence of polarization and partisan bias -- a phenomenon known as the "wisdom of partisan crowds." Generated agents powered by Large Language Models (LLMs) are increasingly used to simulate human collective behavior, yet few benchmarks exist for evaluating their dynamics against the behavior of human groups. In this paper, we examine the extent to which the wisdom of partisan crowds emerges in groups of LLM-based agents that are prompted to role-play as partisan personas (e.g., Democrat or Republican). We find that they not only display human-like partisan biases, but also converge to more accurate beliefs through deliberation as humans do. We then identify several factors that interfere with convergence, including the use of chain-of-thought prompt and lack of details in personas. Conversely, fine-tuning on human data appears to enhance convergence. These findings show the potential and limitations of LLM-based agents as a model of human collective intelligence.

Diganta Misra, Muawiz Chaudhary, Agam Goyal et al.

In an age dominated by resource-intensive foundation models, the ability to efficiently adapt to downstream tasks is crucial. Visual Prompting (VP), drawing inspiration from the prompting techniques employed in Large Language Models (LLMs), has emerged as a pivotal method for transfer learning in the realm of computer vision. As the importance of efficiency continues to rise, research into model compression has become indispensable in alleviating the computational burdens associated with training and deploying over-parameterized neural networks. A primary objective in model compression is to develop sparse and/or quantized models capable of matching or even surpassing the performance of their over-parameterized, full-precision counterparts. Although previous studies have explored the effects of model compression on transfer learning, its impact on visual prompting-based transfer remains unclear. This study aims to bridge this gap, shedding light on the fact that model compression detrimentally impacts the performance of visual prompting-based transfer, particularly evident in scenarios with low data volume. Furthermore, our findings underscore the adverse influence of sparsity on the calibration of downstream visual-prompted models. However, intriguingly, we also illustrate that such negative effects on calibration are not present when models are compressed via quantization. This empirical investigation underscores the need for a nuanced understanding beyond mere accuracy in sparse and quantized settings, thereby paving the way for further exploration in Visual Prompting techniques tailored for sparse and quantized models.

Agam Goyal, Olivia Pal, Hari Sundaram et al.

As autonomous LLM-based agents increasingly populate social platforms, understanding the dynamics of AI-agent communities becomes essential for both communication research and platform governance. We present the first large-scale empirical comparison of AI-agent and human online communities, analyzing 73,899 Moltbook and 189,838 Reddit posts across five matched communities. Structurally, we find that Moltbook exhibits extreme participation inequality (Gini = 0.84 vs. 0.47) and high cross-community author overlap (33.8\% vs. 0.5\%). In terms of linguistic attributes, content generated by AI-agents is emotionally flattened, cognitively shifted toward assertion over exploration, and socially detached. These differences give rise to apparent community-level homogenization, but we show this is primarily a structural artifact of shared authorship. At the author level, individual agents are more identifiable than human users, driven by outlier stylistic profiles amplified by their extreme posting volume. As AI-mediated communication reshapes online discourse, our work offers an empirical foundation for understanding how multi-agent interaction gives rise to collective communication dynamics distinct from those of human communities.

Xianyang Zhan, Agam Goyal, Yilun Chen et al.

Large language models (LLMs) have shown promise in many natural language understanding tasks, including content moderation. However, these models can be expensive to query in real-time and do not allow for a community-specific approach to content moderation. To address these challenges, we explore the use of open-source small language models (SLMs) for community-specific content moderation tasks. We fine-tune and evaluate SLMs (less than 15B parameters) by comparing their performance against much larger open- and closed-sourced models in both a zero-shot and few-shot setting. Using 150K comments from 15 popular Reddit communities, we find that SLMs outperform zero-shot LLMs at content moderation -- 11.5% higher accuracy and 25.7% higher recall on average across all communities. Moreover, few-shot in-context learning leads to only a marginal increase in the performance of LLMs, still lacking compared to SLMs. We further show the promise of cross-community content moderation, which has implications for new communities and the development of cross-platform moderation techniques. Finally, we outline directions for future work on language model based content moderation. Code and models can be found at https://github.com/AGoyal0512/SLM-Mod.

Yian Wang, Yuen Chen, Agam Goyal et al.

Large language models (LLMs) frequently generate toxic content, posing significant risks for safe deployment. Current mitigation strategies often degrade generation quality or require costly human annotation. We propose CAUSALDETOX, a framework that identifies and intervenes on the specific attention heads causally responsible for toxic generation. Using the Probability of Necessity and Sufficiency (PNS), we isolate a minimal set of heads that are necessary and sufficient for toxicity. We utilize these components via two complementary strategies: (1) Local Inference-Time Intervention, which constructs dynamic, input-specific steering vectors for context-aware detoxification, and (2) PNS-Guided Fine-Tuning, which permanently unlearns toxic representations. We also introduce PARATOX, a novel benchmark of aligned toxic/non-toxic sentence pairs enabling controlled counterfactual evaluation. Experiments on ToxiGen, ImplicitHate, and ParaDetox show that CAUSALDETOX achieves up to 5.34% greater toxicity reduction compared to baselines while preserving linguistic fluency, and offers a 7x speedup in head selection.

Agam Goyal, Yian Wang, Eshwar Chandrasekharan et al.

LLM-based social simulations can generate believable community interactions, enabling ``policy wind tunnels'' where governance interventions are tested before deployment. But believability is not causality. Claims like ``intervention $A$ reduces escalation'' require causal semantics that current simulation work typically does not specify. We propose adopting the causal counterfactual framework, distinguishing \textit{necessary causation} (would the outcome have occurred without the intervention?) from \textit{sufficient causation} (does the intervention reliably produce the outcome?). This distinction maps onto different stakeholder needs: moderators diagnosing incidents require evidence about necessity, while platform designers choosing policies require evidence about sufficiency. We formalize this mapping, show how simulation design can support estimation under explicit assumptions, and argue that the resulting quantities should be interpreted as simulator-conditional causal estimates whose policy relevance depends on simulator fidelity. Establishing this framework now is essential: it helps define what adequate fidelity means and moves the field from simulations that look realistic toward simulations that can support policy changes.

Michelle Huang, Agam Goyal, Koustuv Saha et al.

Generative search systems are increasingly replacing link-based retrieval with AI-generated summaries, yet little is known about how these systems differ in sources, language, and fidelity to cited material. We examine responses to 11,000 real search queries across four systems -- vanilla GPT, Search GPT, Google AI Overviews, and traditional Google Search -- at three levels: source diversity, linguistic characterization of the generated summary, and source-summary fidelity. We find that generative search systems exhibit significant \textit{source-selection} biases in their citations, favoring certain sources over others. Incorporating search also selectively attenuates epistemic markers, reducing hedging by up to 60\% while preserving confidence language in the AI-generated summaries. At the same time, AI summaries further compound the citation biases: Wikipedia and longer sources are disproportionately overrepresented, whereas cited social media content and negatively framed sources are substantially underrepresented. Our findings highlight the potential for \textit{answer bubbles}, in which identical queries yield structurally different information realities across systems, with implications for user trust, source visibility, and the transparency of AI-mediated information access.

Olivia Pal, Agam Goyal, Eshwar Chandrasekharan et al.

News consumption on social media has become ubiquitous, yet how different forms of engagement shape psychosocial outcomes remains unclear. To address this gap, we leveraged a large-scale dataset of ~26M posts and ~45M comments on the BlueSky platform, and conducted a quasi-experimental study, matching 81,345 Treated users exposed to News feeds with 83,711 Control users using stratified propensity score analysis. We examined psychosocial wellbeing, in terms of affective, behavioral, and cognitive outcomes. Our findings reveal that news engagement produces systematic trade-offs: increased depression, stress, and anxiety, yet decreased loneliness and increased social interaction on the platform. Regression models reveal that News feed bookmarking is associated with greater psychosocial deterioration compared to commenting or quoting, with magnitude differences exceeding tenfold. These per-engagement effects accumulate with repeated exposure, showing significant psychosocial impacts. Our work extends theories of news effects beyond crisis-centric frameworks by demonstrating that routine consumption creates distinct psychological dynamics depending on engagement type, and bears implications for tools and interventions for mitigating the psychosocial costs of news consumption on social media.

Yian Wang, Agam Goyal, Yuen Chen et al.

LLM agents increasingly rely on persistent state, including transcripts, summaries, retrieved context, and memory buffers, to support long-horizon interaction. This makes safety depend not only on individual model outputs, but also on what an agent stores and later reuses. We study a failure mode we call memory laundering: toxic or adversarial context can be compressed into memory summaries that no longer appear toxic under standard detectors, while still preserving hostile framing or conflict structure that influences future generations. Using paired counterfactual multi-agent rollouts, we show that toxic-origin memory summaries can remain below common toxicity thresholds while nevertheless increasing downstream toxicity relative to matched neutral baselines. To measure this hidden influence, we introduce the sub-threshold propagation gap (SPG), which quantifies downstream behavioral differences conditioned on memory states that a deployed monitor would classify as safe. Our experiments show that toxicity propagates through distinct state channels: raw transcript reuse drives overt downstream toxicity, while compressed memory carries hidden sub-threshold influence. We further find that mitigation depends critically on intervention placement. Sanitizing toxic state before summarization substantially reduces the hidden propagation gap, whereas cleaning only the completed summary can leave laundered influence intact. These results suggest that safety in memory-augmented agents should be treated as a state-control problem over evolving context, with sanitization applied before unsafe information is compressed into persistent memory.

Olivia Pal, Agam Goyal, Eshwar Chandrasekharan et al.

Algorithmic feeds have become primary environments for encountering information online, yet while they shape what people see, less is known about how sustained feed exposure shapes how people write. Drawing on Cultivation Theory, we examine whether algorithmic feeds function as online environments that leave measurable traces in users' language. We leverage a large-scale longitudinal dataset of 235M posts by 4M users on Bluesky, and conduct a quasi-experimental study matching an initial pool of 368,513 users exposed to one of three feeds -- News, Science, and Blacksky -- with a pool of 2,001,915 active control users who did not engage with any of these feeds. We examine linguistic evolution across three dimensions: lexico-semantics, psycholinguistics, and topics. We find that users exposed to these feeds show significantly greater stylistic accommodation, semantic alignment, and register formalization than matched controls. These effects vary markedly by feed identity -- Blacksky produces the deepest psycholinguistic restructuring, with significant shifts in cognitive processing, affective expression, and pronoun use, while News and Science effects are largely confined to register and topical focus. Regression models reveal that reposting is the most consistent predictor of linguistic convergence across all feeds, whereas posting and bookmarking show feed-dependent effects, with effects differing more than fourfold across feeds. Our work extends Cultivation Theory beyond belief formation to linguistic behavior, demonstrating that feeds function as persistent linguistic environments that gradually shape what and how users write online. Our work has implications for studying algorithmic influence, online identity formation, and the design and governance of feed-based platforms that mediate online interactions.

Agam Goyal, Zhaoxing Wu, Richard P. Yim et al.

A system of coupled oscillators on an arbitrary graph is locally driven by the tendency to mutual synchronization between nearby oscillators, but can and often exhibit nonlinear behavior on the whole graph. Understanding such nonlinear behavior has been a key challenge in predicting whether all oscillators in such a system will eventually synchronize. In this paper, we demonstrate that, surprisingly, such nonlinear behavior of coupled oscillators can be effectively linearized in certain latent dynamic spaces. The key insight is that there is a small number of `latent dynamics filters', each with a specific association with synchronizing and non-synchronizing dynamics on subgraphs so that any observed dynamics on subgraphs can be approximated by a suitable linear combination of such elementary dynamic patterns. Taking an ensemble of subgraph-level predictions provides an interpretable predictor for whether the system on the whole graph reaches global synchronization. We propose algorithms based on supervised matrix factorization to learn such latent dynamics filters. We demonstrate that our method performs competitively in synchronization prediction tasks against baselines and black-box classification algorithms, despite its simple and interpretable architecture.

Omkar Gurjar, Agam Goyal, Eshwar Chandrasekharan

Key point extraction is an important task in argument summarization which involves extracting high-level short summaries from arguments. Existing approaches for KP extraction have been mostly evaluated on the popular ArgKP21 dataset. In this paper, we highlight some of the major limitations of the ArgKP21 dataset and demonstrate the need for new benchmarks that are more representative of actual human conversations. Using SoTA large language models (LLMs), we curate a new argument key point extraction dataset called ArgCMV comprising of around 12K arguments from actual online human debates spread across over 3K topics. Our dataset exhibits higher complexity such as longer, co-referencing arguments, higher presence of subjective discourse units, and a larger range of topics over ArgKP21. We show that existing methods do not adapt well to ArgCMV and provide extensive benchmark results by experimenting with existing baselines and latest open source models. This work introduces a novel KP extraction dataset for long-context online discussions, setting the stage for the next generation of LLM-driven summarization research.

Agam Goyal, Xianyang Zhan, Yilun Chen et al.

Large language models (LLMs) have shown great potential in flagging harmful content in online communities. Yet, existing approaches for moderation require a separate model for every community and are opaque in their decision-making, limiting real-world adoption. We introduce Mixture of Moderation Experts (MoMoE), a modular, cross-community framework that adds post-hoc explanations to scalable content moderation. MoMoE orchestrates four operators -- Allocate, Predict, Aggregate, Explain -- and is instantiated as seven community-specialized experts (MoMoE-Community) and five norm-violation experts (MoMoE-NormVio). On 30 unseen subreddits, the best variants obtain Micro-F1 scores of 0.72 and 0.67, respectively, matching or surpassing strong fine-tuned baselines while consistently producing concise and reliable explanations. Although community-specialized experts deliver the highest peak accuracy, norm-violation experts provide steadier performance across domains. These findings show that MoMoE yields scalable, transparent moderation without needing per-community fine-tuning. More broadly, they suggest that lightweight, explainable expert ensembles can guide future NLP and HCI research on trustworthy human-AI governance of online communities.

Agam Goyal, Koyel Mukherjee, Apoorv Saxena et al.

Dense retrievers in retrieval-augmented generation (RAG) systems exhibit systematic biases -- including brevity, position, literal matching, and repetition biases -- that can compromise retrieval quality. Query rewriting techniques are now standard in RAG pipelines, yet their impact on these biases remains unexplored. We present the first systematic study of how query enhancement techniques affect dense retrieval biases, evaluating five methods across six retrievers. Our findings reveal that simple LLM-based rewriting achieves the strongest aggregate bias reduction (54\%), yet fails under adversarial conditions where multiple biases combine. Mechanistic analysis uncovers two distinct mechanisms: simple rewriting reduces bias through increased score variance, while pseudo-document generation methods achieve reduction through genuine decorrelation from bias-inducing features. However, no technique uniformly addresses all biases, and effects vary substantially across retrievers. Our results provide practical guidance for selecting query enhancement strategies based on specific bias vulnerabilities. More broadly, we establish a taxonomy distinguishing query-document interaction biases from document encoding biases, clarifying the limits of query-side interventions for debiasing RAG systems.

Agam Goyal, Vedant Rathi, William Yeh et al.

Large language models (LLMs) are now ubiquitous in user-facing applications, yet they still generate undesirable toxic outputs, including profanity, vulgarity, and derogatory remarks. Although numerous detoxification methods exist, most apply broad, surface-level fixes and can therefore easily be circumvented by jailbreak attacks. In this paper we leverage sparse autoencoders (SAEs) to identify toxicity-related directions in the residual stream of models and perform targeted activation steering using the corresponding decoder vectors. We introduce three tiers of steering aggressiveness and evaluate them on GPT-2 Small and Gemma-2-2B, revealing trade-offs between toxicity reduction and language fluency. At stronger steering strengths, these causal interventions surpass competitive baselines in reducing toxicity by up to 20%, though fluency can degrade noticeably on GPT-2 Small depending on the aggressiveness. Crucially, standard NLP benchmark scores upon steering remain stable, indicating that the model's knowledge and general abilities are preserved. We further show that feature-splitting in wider SAEs hampers safety interventions, underscoring the importance of disentangled feature learning. Our findings highlight both the promise and the current limitations of SAE-based causal interventions for LLM detoxification, further suggesting practical guidelines for safer language-model deployment.

Soorya Ram Shimgekar, Ruining Zhao, Agam Goyal et al.

On social media, many individuals experiencing suicidal ideation (SI) do not disclose their distress explicitly. Instead, signs may surface indirectly through everyday posts or peer interactions. Detecting such implicit signals early is critical but remains challenging. We frame early and implicit SI as a forward-looking prediction task and develop a computational framework that models a user's information environment, consisting of both their longitudinal posting histories as well as the discourse of their socially proximal peers. We adopted a composite network centrality measure to identify top neighbors of a user, and temporally aligned the user's and neighbors' interactions -- integrating the multi-layered signals in a fine-tuned DeBERTa-v3 model. In a Reddit study of 1,000 (500 Case and 500 Control) users, our approach improves early and implicit SI detection by 15% over individual-only baselines. These findings highlight that peer interactions offer valuable predictive signals and carry broader implications for designing early detection systems that capture indirect as well as masked expressions of risk in online environments.

Yun-Shiuan Chuang, Krirk Nirunwiroj, Zach Studdiford et al.

Creating human-like large language model (LLM) agents is crucial for faithful social simulation. Having LLMs role-play based on demographic information sometimes improves human likeness but often does not. This study assessed whether LLM alignment with human behavior can be improved by integrating information from empirically-derived human belief networks. Using data from a human survey, we estimated a belief network encompassing 64 topics loading on nine non-overlapping latent factors. We then seeded LLM-based agents with an opinion on one topic, and assessed the alignment of its expressed opinions on remaining test topics with corresponding human data. Role-playing based on demographic information alone did not align LLM and human opinions, but seeding the agent with a single belief greatly improved alignment for topics related in the belief network, and not for topics outside the network. These results suggest a novel path for human-LLM belief alignment in work seeking to simulate and understand patterns of belief distributions in society.