arXiv subjects

Naman Gupta

Publications and source records attributed to Naman Gupta.

13 recordsLinked to original sources

arXiv2026

Swift Sampling: Selecting Temporal Surprises via Taylor Series

While most frames in long-form video are redundant, the critical information resides in temporal surprises: moments where the actual visual features deviate from their predicted evolution. Inspired by the human brain's predictive coding, we introduce Swift Sampling, an elegant, training-free frame selection algorithm that automatically identifies high-information moments in a video. Specifically, we model a video as a differentiable trajectory in the visual latent space and compute the velocity and acceleration of its features. Then, we apply Taylor expansion to project the expected path of subsequent frames. Frames that diverge sharply from this predicted manifold are identified as temporally surprising frames and selected for sampling. Unlike prior training-free methods that rely on auxiliary networks or video-specific hyperparameter tuning, Swift Sampling is incredibly lightweight, adding only 0.02x additional computational cost over baseline making it 30x cheaper overhead than leading baselines. Across three long-video question answering benchmarks and 10 different downstream tasks, Swift Sampling outperforms uniform sampling and prior query-agnostic baselines. It is especially powerful for long videos with limited frame budgets improving accuracy by up to +12.5 points.

cs.CV↗

arXiv2026

Chow-Liu Ordering for Long-Context Reasoning in Chain-of-Agents

Sequential multi-agent reasoning frameworks such as Chain-of-Agents (CoA) handle long-context queries by decomposing inputs into chunks and processing them sequentially using LLM-based worker agents that read from and update a bounded shared memory. From a probabilistic perspective, CoA aims to approximate the conditional distribution corresponding to a model capable of jointly reasoning over the entire long context. CoA achieves this through a latent-state factorization in which only bounded summaries of previously processed evidence are passed between agents. The resulting bounded-memory approximation introduces a lossy information bottleneck, making the final evidence state inherently dependent on the order in which chunks are processed. In this work, we study the problem of chunk ordering for long-context reasoning. We use the well-known Chow-Liu trees to learn a dependency structure that prioritizes strongly related chunks. Empirically, we show that a breadth-first traversal of the resulting tree yields chunk orderings that reduce information loss across agents and consistently outperform both default document-chunk ordering and semantic score-based ordering in answer relevance and exact-match accuracy across three long-context benchmarks.

cs.CL↗

arXiv2026

A New Dataset and Performance Benchmark for Real-time Spacecraft Segmentation in Onboard Computers

Spacecraft deployed in outer space are routinely subjected to various forms of damage due to exposure to hazardous environments. In addition, there are significant risks to the subsequent process of in-space repairs through human extravehicular activity or robotic manipulation, incurring substantial operational costs. Recent developments in image segmentation could enable the development of reliable and cost-effective autonomous inspection systems. While these models often require large amounts of training data to achieve satisfactory results, publicly available annotated spacecraft segmentation data are very scarce. Here, we present a new dataset of nearly 64k annotated spacecraft images that was created using real spacecraft models, superimposed on a mixture of real and synthetic backgrounds generated using NASA's TTALOS pipeline. To mimic camera distortions and noise in real-world image acquisition, we also added different types of noise and distortion to the images. Our dataset includes images with several real-world challenges, including noise, camera distortions, glare, varying lighting conditions, varying field of view, partial spacecraft visibility, brightly-lit city backgrounds, densely patterned and confounding backgrounds, aurora borealis, and a wide variety of spacecraft geometries. Finally, we finetuned YOLOv8 and YOLOv11 models for spacecraft segmentation to generate performance benchmarks for the dataset under well-defined hardware and inference time constraints to mimic real-world image segmentation challenges for real-time onboard applications in space on NASA's inspector spacecraft. The resulting models, when tested under these constraints, achieved a Dice score of 0.92, Hausdorff distance of 0.69, and an inference time of about 0.5 second. The dataset and models for performance benchmark are available at https://github.com/RiceD2KLab/SWiM.

cs.CV↗

arXiv2026

"Lighting The Way For Those Not Here": How Technology Researchers Can Help Fight the Missing and Murdered Indigenous Relatives (MMIR) Crisis

Indigenous peoples across Turtle Island (North America) face disproportionate rates of disappearance and murder, a "genocide" rooted in settler-colonial violence and systemic erasure. Technology plays a crucial role in the Missing and Murdered Indigenous Relatives (MMIR) crisis: perpetuating harm and impeding investigations, yet enabling advocacy and resistance. Communities utilize technologies such as AMBER alerts, news websites, social media groups, and campaigns (like #MMIW, #MMIWR, #NoMoreStolenSisters, and #NoMoreStolenDaughters) to mobilize searches, amplify awareness, and honor missing relatives. Yet, little research in HCI has critically examined technology's role in shaping the MMIR crisis by centering community voices. Through a large-scale study, we analyze 140 webpages to identify systemic, technological, and institutional barriers that hinder communities' efforts, while highlighting socio-technical actions that foster healing and safety. Finally, we amplify Indigenous voices by providing a dataset of stories that resist epistemic erasure, along with recommendations for HCI researchers to support Indigenous-led initiatives with cultural sensitivity, accountability, and self-determination.

cs.CY↗

arXiv2026

Beyond Polarization: Opinion Mixing and Social Influence in Deliberation

Deliberative processes are often discussed as increasing or decreasing polarization. This approach misses a different, and arguably more diagnostic, dimension of opinion change: whether deliberation reshuffles who agrees with whom, or simply moves everyone in parallel while preserving the pre-deliberation rank ordering. We introduce \opinion mixing, measured by Kendall's rank correlation (τ) between pre- and post-deliberation responses, as a complement to variance-based polarization metrics. Across two large online deliberative polls spanning 32 countries (MCF-2022: n=6,342; MCF-2023: n=1,529), deliberation increases opinion mixing relative to survey-only controls: treatment groups exhibit lower rank correlation on (97%) and (93%) of opinion questions, respectively. Polarization measures based on variance tell a more heterogeneous story: controls consistently converge, while treated groups sometimes converge and sometimes diverge depending on the issue. To probe mechanisms, we link transcripts and surveys in a third event (SOF: (n=617), 116 groups) and use LLM-assisted coding of 6,232 discussion statements. Expressed support in discussion statements strongly predicts subsequent group-level opinion shifts; this correlation is amplified by justification quality in the statements but not by argument novelty. To our knowledge, we are the first to observe how different notions of argument quality have different associations with the outcome of deliberation. This suggests that opinion change after deliberation is related to selective uptake of well-reasoned arguments, producing complex patterns of opinion reorganization that standard polarization metrics may miss.

cs.SI↗

arXiv2025

Formal State-Machine Models for Uniswap v3 Concentrated-Liquidity AMMs: Priced Timed Automata, Finite-State Transducers, and Provable Rounding Bounds

Concentrated-liquidity automated market makers (CLAMMs), as exemplified by Uniswap v3, are now a common primitive in decentralized finance frameworks. Their design combines continuous trading on constant-function curves with discrete tick boundaries at which liquidity positions change and rounding effects accumulate. While there is a body of economic and game-theoretic analysis of CLAMMs, there is negligible work that treats Uniswap v3 at the level of formal state machines amenable to model checking or theorem proving. In this paper we propose a formal modeling approach for Uniswap v3-style CLAMMs using (i) networks of priced timed automata (PTA), and (ii) finite-state transducers (FST) over discrete ticks. Positions are treated as stateful objects that transition only when the pool price crosses the ticks that bound their active range. We show how to encode the piecewise constant-product invariant, fee-growth variables, and tick-crossing rules in a PTA suitable for tools such as UPPAAL, and how to derive a tick-level FST abstraction for specification in TLA+. We define an explicit tick-wise invariant for a discretized, single-tick CLAMM model and prove that it is preserved up to a tight additive rounding bound under fee-free swaps. This provides a formal justification for the "$ε$-slack" used in invariance properties and shows how rounding enters as a controlled perturbation. We then instantiate these models in TLA+ and use TLC to exhaustively check the resulting invariants on structurally faithful instances, including a three-tick concentrated-liquidity configuration and a bounded no-rounding-only-arbitrage property in a bidirectional single-tick model. We discuss how these constructions lift to the tick-wise structure of Uniswap v3 via virtual reserves, and how the resulting properties can be phrased as PTA/TLA+ invariants about cross-tick behaviour and rounding safety.

cs.LO↗

arXiv2025

STACKFEED: Structured Textual Actor-Critic Knowledge Base Editing with FeedBack

Large Language Models (LLMs) often generate incorrect or outdated information, especially in low-resource settings or when dealing with private data. To address this, Retrieval-Augmented Generation (RAG) uses external knowledge bases (KBs), but these can also suffer from inaccuracies. We introduce STACKFEED, a novel Structured Textual Actor-Critic Knowledge base editing with FEEDback approach that iteratively refines the KB based on expert feedback using a multi-actor, centralized critic reinforcement learning framework. STACKFEED defines a ReACT actor agent on each document to perform structured edits based on document specific targeted instructions. Experimental results showcase that STACKFEED significantly improves KB quality and performance of the RAG system. We evaluate STACKFEED on low-resource programming problems, modified python packaged and factual question-answering tasks.

cs.AI↗

arXiv2025

FABRIC: Framework for Agent-Based Realistic Intelligence Creation

Large language models (LLMs) are increasingly deployed as agents, expected to decompose goals, invoke tools, and verify results in dynamic environments. Realizing these capabilities requires access to agentic data-structured interaction records that couple user intents with tool specifications, argument-grounded calls, and verifiable execution traces. However, collecting such data from human annotators is costly, time-consuming, and difficult to scale. We present a unified framework for synthesizing agentic data using only LLMs, without any human-in-the-loop supervision. This framework decomposes generation into modular pipelines that produce complete interaction records spanning task specifications, tool definitions, policy pseudocode, natural language exchanges, and execution traces. Records conform to strict syntactic and semantic constraints, ensuring machine-parseability and faithful alignment across inputs, outputs, and tool calls. Beyond single tasks, there is support for both multi-task and multi-turn agent interactions, enabling the construction of datasets that reflect the full spectrum of tool-use competencies. To ensure quality and consistency, the framework integrates constrained generation formats, JSON-schema validation, and judge-based filtering. This paper formalizes the schema for agentic records, details the prompt design principles that guide generation, and introduces scalable pipelines for high-quality synthetic data. By providing a reproducible, LLM-only alternative to manual collection, hence advancing the development of agentic LLMs capable of robust tool use.

cs.AI↗

arXiv2025

COSMIR: Chain Orchestrated Structured Memory for Iterative Reasoning over Long Context

Reasoning over very long inputs remains difficult for large language models (LLMs). Common workarounds either shrink the input via retrieval (risking missed evidence), enlarge the context window (straining selectivity), or stage multiple agents to read in pieces. In staged pipelines (e.g., Chain of Agents, CoA), free-form summaries passed between agents can discard crucial details and amplify early mistakes. We introduce COSMIR (Chain Orchestrated Structured Memory for Iterative Reasoning), a chain-style framework that replaces ad hoc messages with a structured memory. A Planner agent first turns a user query into concrete, checkable sub-questions. worker agents process chunks via a fixed micro-cycle: Extract, Infer, Refine, writing all updates to the shared memory. A Manager agent then Synthesizes the final answer directly from the memory. This preserves step-wise read-then-reason benefits while changing both the communication medium (structured memory) and the worker procedure (fixed micro-cycle), yielding higher faithfulness, better long-range aggregation, and auditability. On long-context QA from the HELMET suite, COSMIR reduces propagation-stage information loss and improves accuracy over a CoA baseline.

cs.AI↗

arXiv2022

Generalized Product-of-Experts for Learning Multimodal Representations in Noisy Environments

A real-world application or setting involves interaction between different modalities (e.g., video, speech, text). In order to process the multimodal information automatically and use it for an end application, Multimodal Representation Learning (MRL) has emerged as an active area of research in recent times. MRL involves learning reliable and robust representations of information from heterogeneous sources and fusing them. However, in practice, the data acquired from different sources are typically noisy. In some extreme cases, a noise of large magnitude can completely alter the semantics of the data leading to inconsistencies in the parallel multimodal data. In this paper, we propose a novel method for multimodal representation learning in a noisy environment via the generalized product of experts technique. In the proposed method, we train a separate network for each modality to assess the credibility of information coming from that modality, and subsequently, the contribution from each modality is dynamically varied while estimating the joint distribution. We evaluate our method on two challenging benchmarks from two diverse domains: multimodal 3D hand-pose estimation and multimodal surgical video segmentation. We attain state-of-the-art performance on both benchmarks. Our extensive quantitative and qualitative evaluations show the advantages of our method compared to previous approaches.

cs.CV↗

arXiv2021

TradeChain: Decoupling Traceability and Identity inBlockchain enabled Supply Chains

In this work, we propose a privacy-preservation framework, TradeChain, which decouples the trade events of participants using decentralised identities. TradeChain adopts the Self-Sovereign Identity (SSI) principles and makes the following novel contributions: a) it incorporates two separate ledgers: a public permissioned blockchain for maintaining identities and the permissioned blockchain for recording trade flows, b) it uses Zero Knowledge Proofs (ZKPs) on traders' private credentials to prove multiple identities on trade ledger and c) allows data owners to define dynamic access rules for verifying traceability information from the trade ledger using access tokens and Ciphertext Policy Attribute-Based Encryption (CP-ABE). A proof of concept implementation of TradeChain is presented on Hyperledger Indy and Fabric and an extensive evaluation of execution time, latency and throughput reveals minimal overheads.

cs.CR↗

arXiv2017

Fastron: An Online Learning-Based Model and Active Learning Strategy for Proxy Collision Detection

We introduce the Fastron, a configuration space (C-space) model to be used as a proxy to kinematic-based collision detection. The Fastron allows iterative updates to account for a changing environment through a combination of a novel formulation of the kernel perceptron learning algorithm and an active learning strategy. Our simulations on a 7 degree-of-freedom arm indicate that proxy collision checks may be performed at least 2 times faster than an efficient polyhedral collision checker and at least 8 times faster than an efficient high-precision collision checker. The Fastron model provides conservative collision status predictions by padding C-space obstacles, and proxy collision checking time does not scale poorly as the number of workspace obstacles increases. All results were achieved without GPU acceleration or parallel computing.

cs.RO↗

arXiv2017

Trajectory Tracking Using Motion Primitives for the Purcell's Swimmer

Locomotion at low Reynolds numbers is a topic of growing interest, spurred by its various engineering and medical applications. This paper presents a novel prototype and a locomotion algorithm for the 3-link planar Purcell's swimmer based on Lie algebraic notions. The kinematic model based on Cox theory of the prototype swimmer is a driftless control-affine system. Using the existing strong controllability and related results, the existence of motion primitives is initially shown. The Lie algebra of the control vector fields is then used to synthesize control profiles to generate motions along the basis of the Lie algebra associated with the structure group of the system. An open loop control system with vision-based positioning is successfully implemented which allows tracking any given continuous trajectory of the position and orientation of the swimmer's base link. Alongside, the paper also provides a theoretical interpretation of the symmetry arguments presented in the existing literature to generate the control profiles of the swimmer.

eess.SY↗