AIConfPaper
← Search

Arjun Guha

8 accepted papers

2026

Agnostics: Learning to Synthesize Code in Any Programming Language with a Universal Reinforcement Learning Environment

Aleksander Boruch-Gruszecki, Yangtian Zi, Zixuan Wu, Tejas Oberoi, Carolyn Jane Anderson, Joydeep Biswas +1
ICLR 2026poster

Large language models (LLMs) already excel at writing code in high-resource languages such as Python and JavaScript, yet stumble on low-resource languages that remain essential to science and engineering. Besides the obvious shortage of pre-training data, post-training itself is a bottleneck: every…

Cited by 0Source ↗code ↗Scholar ↗
2025

NNsight and NDIF: Democratizing Access to Open-Weight Foundation Model Internals

Jaden Fried Fiotto-Kaufman, Alexander Russell Loftus, Eric Todd, Jannik Brinkmann, Koyena Pal, Dmitrii Troitskii +14
ICLR 2025poster

We introduce NNsight and NDIF, technologies that work in tandem to enable scientific study of the representations and computations learned by very large neural networks. NNsight is an open-source system that extends PyTorch to introduce deferred remote execution. The National Deep Inference Fabric (…

Cited by 2Source ↗PDF ↗code ↗Scholar ↗
2025

Substance Beats Style: Why Beginning Students Fail to Code with LLMs

Francesca Lucchetti, Zixuan Wu, Arjun Guha, Molly Q Feldman, Carolyn Jane Anderson
NAACL 2025long

Although LLMs are increasing the productivity of professional programmers, existing work shows that beginners struggle to prompt LLMs to solve text-to-code tasks (Nguyen et al., 2024; Prather et al., 2024b; Mordechai et al., 2024). Why is this the case? This paper explores two competing hypotheses a…

Cited by 3Source ↗PDF ↗code ↗Scholar ↗
2024

Deploying and Evaluating LLMs to Program Service Mobile Robots

Zichao Hu, Francesca Lucchetti, Claire Schlesinger, Yash Saxena, Anders Freeman, Sadanand Modak +2
RA-L 2024

Recent advancements in large language models (LLMs) have spurred interest in using them for generating robot programs from natural language, with promising initial results. We investigate the use of LLMs to generate programs for service mobile robots leveraging mobility, perception, and human intera

Cited by 60Source ↗code ↗Scholar ↗
2024

SelfCodeAlign: Self-Alignment for Code Generation

Yuxiang Wei, Federico Cassano, Jiawei Liu, Yifeng Ding, Naman Jain, Zachary Mueller +4
NeurIPS 2024poster

Instruction tuning is a supervised fine-tuning approach that significantly improves the ability of large language models (LLMs) to follow human instructions. For programming tasks, most models are finetuned with costly human-annotated instruction-response pairs or those generated by large, proprieta…

Cited by 3Source ↗PDF ↗code ↗Scholar ↗
2024

StudentEval: A Benchmark of Student-Written Prompts for Large Language Models of Code

Hannah Babe, Sydney Nguyen, Yangtian Zi, Arjun Guha, Molly Feldman, Carolyn Anderson
ACL 2024findings

Code LLMs have the potential to make it easier for non-experts to understand and write code. However, current CodeLLM benchmarks rely on a single expert-written prompt per problem, making it hard to generalize their success to non-expert users. In this paper, we present a new natural-language-to-cod…

Cited by 36Source ↗PDF ↗code ↗Scholar ↗
2021

Iterative Program Synthesis for Adaptable Social Navigation

Jarrett Holtz, Simon Andrews, Arjun Guha, Joydeep Biswas
IROS 2021poster

Robot social navigation is influenced by human preferences and environment-specific scenarios such as elevators and doors, thus necessitating end-user adaptability. State-of-the-art approaches to social navigation fall into two categories: model-based social constraints and learning-based approaches…

Cited by 9Source ↗code ↗Scholar ↗
2020

Robot Action Selection Learning via Layered Dimension Informed Program Synthesis

Jarrett Holtz, Arjun Guha, Joydeep Biswas
CoRL 2020

Abstract: Action selection policies (ASPs), used to compose low-level robot skills into complex high-level tasks are commonly represented as neural networks (NNs) in the state of the art. Such a paradigm, while very effective, suffers from a few key problems: 1) NNs are opaque to the user and hence

Cited by 0Source ↗PDF ↗Scholar ↗