arXiv subjects

Navid Madani

Publications and source records attributed to Navid Madani.

7 recordsLinked to original sources

arXiv2026

DUET -- Dual User Embedding Transformers for Offsite Conversion Prediction

Offsite conversion rate (OCVR) prediction is an important ranking problem in computational recommendation systems. This task presents a modeling challenge: click signals are abundant and exhibit short temporal horizons, whereas conversion signals are inherently sparse, long-delayed, and frequently unattributed. Despite these statistical disparities, both signal types must inform models that operate within strict serving-latency constraints. Prior pre-training approaches address this heterogeneity with a single, undifferentiated encoder applied uniformly across both data streams. We propose DUET (Dual User Embedding Transformers), a framework that explicitly partitions user behavioral data into two domain-coherent streams -- clicks and conversions -- and pre-trains dedicated transformer encoders with architectures tailored to each stream's statistical characteristics: multi-layer self-attention for the dense click stream and interleaved cross- and self-attention for the sparse conversion stream. The resulting complementary embeddings are jointly consumed by a downstream ranker without exceeding serving-latency budgets. Evaluation demonstrates up to 0.38% normalized entropy (NE) reduction relative to the strongest baseline, and A/B test shows consistent improvements in OCVR prediction accuracy.

cs.LG↗

arXiv2025

ESC-Judge: A Framework for Comparing Emotional Support Conversational Agents

Large language models (LLMs) increasingly power mental-health chatbots, yet the field still lacks a scalable, theory-grounded way to decide which model is most effective to deploy. We present ESC-Judge, the first end-to-end evaluation framework that (i) grounds head-to-head comparisons of emotional-support LLMs in Clara Hill's established Exploration-Insight-Action counseling model, providing a structured and interpretable view of performance, and (ii) fully automates the evaluation pipeline at scale. ESC-Judge operates in three stages: first, it synthesizes realistic help-seeker roles by sampling empirically salient attributes such as stressors, personality, and life history; second, it has two candidate support agents conduct separate sessions with the same role, isolating model-specific strategies; and third, it asks a specialized judge LLM to express pairwise preferences across rubric-anchored skills that span the Exploration, Insight, and Action spectrum. In our study, ESC-Judge matched PhD-level annotators on 85 percent of Exploration, 83 percent of Insight, and 86 percent of Action decisions, demonstrating human-level reliability at a fraction of the cost. All code, prompts, synthetic roles, transcripts, and judgment scripts are released to promote transparent progress in emotionally supportive AI.

cs.CL↗

arXiv2024

A Recipe For Building a Compliant Real Estate Chatbot

In recent years, there has been significant effort to align large language models with human preferences. This work focuses on developing a chatbot specialized in the real estate domain, with an emphasis on incorporating compliant behavior to ensure it can be used without perpetuating discriminatory practices like steering and redlining, which have historically plagued the real estate industry in the United States. Building on prior work, we present a method for generating a synthetic general instruction-following dataset, along with safety data. Through extensive evaluations and benchmarks, we fine-tuned a llama-3-8B-instruct model and demonstrated that we can enhance it's performance significantly to match huge closed-source models like GPT-4o while making it safer and more compliant. We open-source the model, data and code to support further development and research in the community.

cs.CL↗

arXiv2024

Measuring Dimensions of Self-Presentation in Twitter Bios and their Links to Misinformation Sharing

Social media platforms provide users with a profile description field, commonly known as a ``bio," where they can present themselves to the world. A growing literature shows that text in these bios can improve our understanding of online self-presentation and behavior, but existing work relies exclusively on keyword-based approaches to do so. We here propose and evaluate a suite of \hl{simple, effective, and theoretically motivated} approaches to embed bios in spaces that capture salient dimensions of social meaning, such as age and partisanship. We \hl{evaluate our methods on four tasks, showing that the strongest one out-performs several practical baselines.} We then show the utility of our method in helping understand associations between self-presentation and the sharing of URLs from low-quality news sites on Twitter\hl{, with a particular focus on explore the interactions between age and partisanship, and exploring the effects of self-presentations of religiosity}. Our work provides new tools to help computational social scientists make use of information in bios, and provides new insights into how misinformation sharing may be perceived on Twitter.

cs.CL↗

arXiv2024

Steering Conversational Large Language Models for Long Emotional Support Conversations

In this study, we address the challenge of enabling large language models (LLMs) to consistently adhere to emotional support strategies in extended conversations. We focus on the steerability of the Llama-2 and Llama-3 suite of models, examining their ability to maintain these strategies throughout interactions. To assess this, we introduce the Strategy Relevant Attention (SRA) metric, which quantifies the model's adherence to the prompted strategy through attention maps. To facilitate our study, we create a strategy-conditioned synthetic conversational dataset derived from the ESConv dataset. We also propose various baselines informed by our proposed SRA metric to address the challenge and propose a fine-tuned model that significantly enhances the steerability of the base model in following the strategy throughout the conversation. The code and data are publicly available on our GitHub.

cs.CL↗

arXiv2023

Domain Specific Question Answering Over Knowledge Graphs Using Logical Programming and Large Language Models

Answering questions over domain-specific graphs requires a tailored approach due to the limited number of relations and the specific nature of the domain. Our approach integrates classic logical programming languages into large language models (LLMs), enabling the utilization of logical reasoning capabilities to tackle the KGQA task. By representing the questions as Prolog queries, which are readable and near close to natural language in representation, we facilitate the generation of programmatically derived answers. To validate the effectiveness of our approach, we evaluate it using a well-known benchmark dataset, MetaQA. Our experimental results demonstrate that our method achieves accurate identification of correct answer entities for all test questions, even when trained on a small fraction of annotated data. Overall, our work presents a promising approach to addressing question answering over domain-specific graphs, offering an explainable and robust solution by incorporating logical programming languages.

cs.LG↗

arXiv2021

A Method to Analyze Multiple Social Identities in Twitter Bios

Twitter users signal social identity in their profile descriptions, or bios, in a number of important but complex ways that are not well-captured by existing characterizations of how identity is expressed in language. Better ways of defining and measuring these expressions may therefore be useful both in understanding how social identity is expressed in text, and how the self is presented on Twitter. To this end, the present work makes three contributions. First, using qualitative methods, we identify and define the concept of a personal identifier, which is more representative of the ways in which identity is signaled in Twitter bios. Second, we propose a method to extract all personal identifiers expressed in a given bio. Finally, we present a series of validation analyses that explore the strengths and limitations of our proposed method. Our work opens up exciting new opportunities at the intersection between the social psychological study of social identity and the study of how we compose the self through markers of identity on Twitter and in social media more generally.

cs.SI↗