Showing 1–7 of 7 results for author: Mahale, A

Explanation Multiplicity: Circuit-Level Interpretability Evidence Does Not Survive Defensible Analytic Variation

Authors: Ajay Pravin Mahale

Abstract: The EU AI Act requires providers of high-risk systems to file technical documentation describing how the system reaches its decisions. Mechanistic interpretability is the obvious source of such evidence, and circuit discovery is its most developed instrument. We ask whether that evidence survives the condition under which it would be relied upon: two competent analysts, the same system, the same t… ▽ More The EU AI Act requires providers of high-risk systems to file technical documentation describing how the system reaches its decisions. Mechanistic interpretability is the obvious source of such evidence, and circuit discovery is its most developed instrument. We ask whether that evidence survives the condition under which it would be relied upon: two competent analysts, the same system, the same tool, different defensible settings. We pre-registered a crossed grid of seven analytic axes, every level taken from a published implementation, and mapped each discovered circuit through a deterministic claim map to a structured Annex IV statement. Across 15,840 pre-registered specifications on GPT-2 small and the indirect object identification task, of which 7,561 produced a claim, the derived statement flips across 73.2% of specification pairs (95% CI 0.725 to 0.738) and the modal claim commands 41.1% of the space. The evidence fails a filability criterion at every tolerance a conformity assessment body would plausibly accept. Standardising the single most influential choice, the evaluation metric, leaves the flip rate at 59.4%. Removing circuit size from the claim entirely and holding it fixed leaves 27.1% (95% CI 0.255 to 0.286), still above the pre-registered threshold. The circuits underlying these claims are structurally near-disjoint, median pairwise Jaccard overlap 4%, and functionally uncorrelated at Cohen's kappa 0.015, so the instability is not one mechanism described in different words. We give the filability criterion as a standalone protocol, and we report that one of the seven documented discovery objectives does not execute at all on the library's own canonical task. The study covers one model and one task, and whether the conclusion holds at scale is untested. △ Less

Submitted 13 August, 2026; originally announced August 2026.

Comments: 12 pages, 1 figure, 7 tables. Pre-registered analysis plan; code and results available

Causally Grounded Mechanistic Interpretability for LLMs with Faithful Natural-Language Explanations

Authors: Ajay Pravin Mahale

Abstract: Mechanistic interpretability identifies internal circuits responsible for model behaviors, yet translating these findings into human-understandable explanations remains an open problem. We present a pipeline that bridges circuit-level analysis and natural language explanations by (i) identifying causally important attention heads via activation patching, (ii) generating explanations using both tem… ▽ More Mechanistic interpretability identifies internal circuits responsible for model behaviors, yet translating these findings into human-understandable explanations remains an open problem. We present a pipeline that bridges circuit-level analysis and natural language explanations by (i) identifying causally important attention heads via activation patching, (ii) generating explanations using both template-based and LLM-based methods, and (iii) evaluating faithfulness using ERASER-style metrics adapted for circuit-level attribution. We evaluate on the Indirect Object Identification (IOI) task in GPT-2 Small (124M parameters), identifying six attention heads accounting for 61.4% of the logit difference. Our circuit-based explanations achieve 100% sufficiency but only 22% comprehensiveness, revealing distributed backup mechanisms. LLM-generated explanations outperform template baselines by 64% on quality metrics. We find no correlation (r = 0.009) between model confidence and explanation faithfulness, and identify three failure categories explaining when explanations diverge from mechanisms. △ Less

Submitted 12 February, 2026; originally announced March 2026.

Comments: 8 pages, 7 figures, 4 tables. MSc thesis work conducted at Hochschule Trier (2026). Code will be released upon publication

MSC Class: 68T50 ACM Class: I.2.7; I.2.6

Learning to Call: A Field Trial of a Collaborative Bandit Algorithm for Improved Message Delivery in Mobile Maternal Health

Authors: Arpan Dasgupta, Mizhaan Maniyar, Awadhesh Srivastava, Sanat Kumar, Amrita Mahale, Aparna Hegde, Arun Suggala, Karthikeyan Shanmugam, Aparna Taneja, Milind Tambe

Abstract: Mobile health (mHealth) programs utilize automated voice messages to deliver health information, particularly targeting underserved communities, demonstrating the effectiveness of using mobile technology to disseminate crucial health information to these populations, improving health outcomes through increased awareness and behavioral change. India's Kilkari program delivers vital maternal health… ▽ More Mobile health (mHealth) programs utilize automated voice messages to deliver health information, particularly targeting underserved communities, demonstrating the effectiveness of using mobile technology to disseminate crucial health information to these populations, improving health outcomes through increased awareness and behavioral change. India's Kilkari program delivers vital maternal health information via weekly voice calls to millions of mothers. However, the current random call scheduling often results in missed calls and reduced message delivery. This study presents a field trial of a collaborative bandit algorithm designed to optimize call timing by learning individual mothers' preferred call times. We deployed the algorithm with around $6500$ Kilkari participants as a pilot study, comparing its performance to the baseline random calling approach. Our results demonstrate a statistically significant improvement in call pick-up rates with the bandit algorithm, indicating its potential to enhance message delivery and impact millions of mothers across India. This research highlights the efficacy of personalized scheduling in mobile health interventions and underscores the potential of machine learning to improve maternal health outreach at scale. △ Less

Submitted 24 November, 2025; v1 submitted 22 July, 2025; originally announced July 2025.

Improving Health Information Access in the World's Largest Maternal Mobile Health Program via Bandit Algorithms

Authors: Arshika Lalan, Shresth Verma, Paula Rodriguez Diaz, Panayiotis Danassis, Amrita Mahale, Kumar Madhu Sudan, Aparna Hegde, Milind Tambe, Aparna Taneja

Abstract: Harnessing the wide-spread availability of cell phones, many nonprofits have launched mobile health (mHealth) programs to deliver information via voice or text to beneficiaries in underserved communities, with maternal and infant health being a key area of such mHealth programs. Unfortunately, dwindling listenership is a major challenge, requiring targeted interventions using limited resources. Th… ▽ More Harnessing the wide-spread availability of cell phones, many nonprofits have launched mobile health (mHealth) programs to deliver information via voice or text to beneficiaries in underserved communities, with maternal and infant health being a key area of such mHealth programs. Unfortunately, dwindling listenership is a major challenge, requiring targeted interventions using limited resources. This paper focuses on Kilkari, the world's largest mHealth program for maternal and child care - with over 3 million active subscribers at a time - launched by India's Ministry of Health and Family Welfare (MoHFW) and run by the non-profit ARRMAN. We present a system called CHAHAK that aims to reduce automated dropouts as well as boost engagement with the program through the strategic allocation of interventions to beneficiaries. Past work in a similar domain has focused on a much smaller scale mHealth program and used markovian restless multiarmed bandits to optimize a single limited intervention resource. However this paper demonstrates the challenges in adopting a markovian approach in Kilkari; therefore CHAHAK instead relies on non-markovian time-series restless bandits, and optimizes multiple interventions to improve listenership. We use real Kilkari data from the Odisha state in India to show CHAHAK's effectiveness in harnessing multiple interventions to boost listenership, benefiting marginalized communities. When deployed CHAHAK will assist the largest maternal mHealth program to date. △ Less

Submitted 14 May, 2024; originally announced July 2024.

Comments: Published at Innovative Applications of Artificial Intelligence (IAAI 2024)

Analyzing and Predicting Low-Listenership Trends in a Large-Scale Mobile Health Program: A Preliminary Investigation

Authors: Arshika Lalan, Shresth Verma, Kumar Madhu Sudan, Amrita Mahale, Aparna Hegde, Milind Tambe, Aparna Taneja

Abstract: Mobile health programs are becoming an increasingly popular medium for dissemination of health information among beneficiaries in less privileged communities. Kilkari is one of the world's largest mobile health programs which delivers time sensitive audio-messages to pregnant women and new mothers. We have been collaborating with ARMMAN, a non-profit in India which operates the Kilkari program, to… ▽ More Mobile health programs are becoming an increasingly popular medium for dissemination of health information among beneficiaries in less privileged communities. Kilkari is one of the world's largest mobile health programs which delivers time sensitive audio-messages to pregnant women and new mothers. We have been collaborating with ARMMAN, a non-profit in India which operates the Kilkari program, to identify bottlenecks to improve the efficiency of the program. In particular, we provide an initial analysis of the trajectories of beneficiaries' interaction with the mHealth program and examine elements of the program that can be potentially enhanced to boost its success. We cluster the cohort into different buckets based on listenership so as to analyze listenership patterns for each group that could help boost program success. We also demonstrate preliminary results on using historical data in a time-series prediction to identify beneficiary dropouts and enable NGOs in devising timely interventions to strengthen beneficiary retention. △ Less

Submitted 13 November, 2023; originally announced November 2023.

Comments: Accepted to Data Science for Social Good Workshop, KDD 2023

Impact of data-splits on generalization: Identifying COVID-19 from cough and context

Authors: Makkunda Sharma, Nikhil Shenoy, Jigar Doshi, Piyush Bagad, Aman Dalmia, Parag Bhamare, Amrita Mahale, Saurabh Rane, Neeraj Agrawal, Rahul Panicker

Abstract: Rapidly scaling screening, testing and quarantine has shown to be an effective strategy to combat the COVID-19 pandemic. We consider the application of deep learning techniques to distinguish individuals with COVID from non-COVID by using data acquirable from a phone. Using cough and context (symptoms and meta-data) represent such a promising approach. Several independent works in this direction h… ▽ More Rapidly scaling screening, testing and quarantine has shown to be an effective strategy to combat the COVID-19 pandemic. We consider the application of deep learning techniques to distinguish individuals with COVID from non-COVID by using data acquirable from a phone. Using cough and context (symptoms and meta-data) represent such a promising approach. Several independent works in this direction have shown promising results. However, none of them report performance across clinically relevant data splits. Specifically, the performance where the development and test sets are split in time (retrospective validation) and across sites (broad validation). Although there is meaningful generalization across these splits the performance significantly varies (up to 0.1 AUC score). In addition, we study the performance of symptomatic and asymptomatic individuals across these three splits. Finally, we show that our model focuses on meaningful features of the input, cough bouts for cough and relevant symptoms for context. The code and checkpoints are available at https://github.com/WadhwaniAI/cough-against-covid △ Less

Submitted 5 June, 2021; originally announced June 2021.

Comments: Published as a workshop paper at ICLR 2021 AI for Public Health Workshop and ICLR 20201 Machine Learning for Preventing and Combating Pandemics Workshop

Cough Against COVID: Evidence of COVID-19 Signature in Cough Sounds

Authors: Piyush Bagad, Aman Dalmia, Jigar Doshi, Arsha Nagrani, Parag Bhamare, Amrita Mahale, Saurabh Rane, Neeraj Agarwal, Rahul Panicker

Abstract: Testing capacity for COVID-19 remains a challenge globally due to the lack of adequate supplies, trained personnel, and sample-processing equipment. These problems are even more acute in rural and underdeveloped regions. We demonstrate that solicited-cough sounds collected over a phone, when analysed by our AI model, have statistically significant signal indicative of COVID-19 status (AUC 0.72, t-… ▽ More Testing capacity for COVID-19 remains a challenge globally due to the lack of adequate supplies, trained personnel, and sample-processing equipment. These problems are even more acute in rural and underdeveloped regions. We demonstrate that solicited-cough sounds collected over a phone, when analysed by our AI model, have statistically significant signal indicative of COVID-19 status (AUC 0.72, t-test,p <0.01,95% CI 0.61-0.83). This holds true for asymptomatic patients as well. Towards this, we collect the largest known(to date) dataset of microbiologically confirmed COVID-19 cough sounds from 3,621 individuals. When used in a triaging step within an overall testing protocol, by enabling risk-stratification of individuals before confirmatory tests, our tool can increase the testing capacity of a healthcare system by 43% at disease prevalence of 5%, without additional supplies, trained personnel, or physical infrastructure △ Less

Submitted 23 September, 2020; v1 submitted 17 September, 2020; originally announced September 2020.

Comments: Under submission to AAAI 20