Aman Sharma

Aman Sharma

MS CS @ Georgia Tech · LLM reasoning & agent systems

I'm an MS CS student and graduate student researcher at Georgia Tech, advised by Prof. Chao Zhang. I study how AI systems reason, adapt to unfamiliar domains, and coordinate, with interests in continual learning, reinforcement learning, and efficient LLM systems.

Two of my papers were accepted at NeurIPS 2026. Frontier Coding Agents Use Metaprogramming was accepted to the Main Track, and EsoLang-Bench to the Evaluations and Datasets Track. Together, they explore how coding agents handle unfamiliar programming languages and build tools to improve their performance.

Previously, I was a research engineer at Lossfunk AI Lab, working on LLM evaluation, metaprogramming, and efficient reasoning. I also worked with Ayush Chopra at the MIT Media Lab on large population models and multi-agent coordination, including the Ripple Effect Protocol developed with MIT and Cisco.

Earlier, I was a Google Summer of Code fellow at NumFOCUS (implementing automatic differentiation rules for image processing in Julia), an ML engineer at Ikigai Labs (benchmarking LLMs vs traditional ML), and an MLH Fellow building high-performance Rust systems for Solana. I won 12+ hackathons and conducted ML workshops and teaching sessions for 200+ undergraduate students.

Resume PDF · September 2026

Research Engineer

Research Highlights

Coding Agents & Evaluation

Multi-Agent Systems

LLM Reasoning & Efficient Systems

My World of Thoughts

all posts →

My World of Research

First page of Frontier Coding Agents Use Metaprogramming to Adapt to Unfamiliar Programming Languages
NeurIPS '26
Frontier Coding Agents Use Metaprogramming to Adapt to Unfamiliar Programming Languages
Aman Sharma, Sushrut Thorat, Paras Chopra
NeurIPS 2026 Main Track · Accepted as a poster
We study how frontier coding agents adapt to unfamiliar programming languages through metaprogramming. On EsoLang-Bench's Brainfuck tasks, reusable Python helpers combined with guidance raised Sonnet 4.6's solve rate from 15% to 80%, demonstrating how agents can build tools to support adaptation to new coding domains.
arXiv
First page of EsoLang-Bench
NeurIPS '26
EsoLang-Bench: Evaluating Large Language Models via Esoteric Programming Languages
Aman Sharma, Chopra, P.
NeurIPS 2026 Evaluations and Datasets Track · Accepted as a poster
We introduce EsoLang-Bench, a benchmark that evaluates genuine reasoning in LLMs using esoteric programming languages where training data is virtually nonexistent. Five frontier models scored 85-95% on standard benchmarks but achieved only 11.2% maximum on EsoLang-Bench, with most models scoring below 5%. All models scored exactly 0% beyond the "Easy" difficulty tier, suggesting fundamental reasoning limitations rather than gradual degradation.
arXiv Website
First page of Ripple Effect Protocol
Project Iceberg
Ripple Effect Protocol: Coordinating Agent Populations
Chopra, A., Sharma, A., Ahmad, F., Muscariello, L., Pandey, V., Raskar, R.
Project Iceberg
Modern AI agents can exchange messages using protocols such as A2A and ACP, yet these mechanisms emphasize communication over coordination. We introduce the Ripple Effect Protocol (REP), where agents share not only their decisions but also lightweight sensitivities — signals expressing how their choices would change if key environmental variables shifted. REP improves coordination accuracy and efficiency over A2A by 41-100% across supply chain, scheduling, and resource allocation benchmarks.
arXiv
First-page excerpt of Do Language Models Deceive?
ICLR '26 WS
Do Language Models Deceive? Strategic Behavior and Emergent Deception in Multi-Agent Auctions
Aman Sharma
ICLR 2026 Workshop - MALGAI
We present the first systematic study of LLM behavior in competitive auction settings. LLMs engage in deceptive behavior in 44% of competitive interactions without explicit instruction, self-classifying tactics such as false disinterest and strategic misdirection while maintaining divergent public and private reasoning. Models also systematically undervalue artwork without provenance metadata and accurately detect AI-generated artwork without labels, demonstrating that competitive contexts induce strategic behavior diverging from stated intentions.
OpenReview
First page of The Sequential Edge
NeurIPS '25 WS
The Sequential Edge: Inverse-Entropy Voting Beats Parallel Self-Consistency at Matched Compute
Aman Sharma, Chopra, P.
NeurIPS 2025 - Efficient Reasoning Workshop
We show that sequential reasoning outperforms parallel self-consistency in 95.6% of configurations at matched compute, with accuracy gains up to 46.7%. On AIME-2025 with Qwen3-235B, sequential scaling achieves 76.7% vs parallel's 30.0%. We introduce inverse-entropy weighted voting, a training-free aggregation method that achieved optimal performance in 97% of sequential runs by downweighting high-entropy (uncertain) outputs.
arXiv Workshop
First page of Think Just Enough
NeurIPS '25 WS
Think Just Enough: Sequence-Level Entropy as a Confidence Signal for LLM Reasoning
Aman Sharma, Chopra, P.
NeurIPS 2025 - FoRLM Workshop
We demonstrate that post-trained models can recognize correct solutions through output entropy analysis. Sequence-level entropy cleanly separates correct from incorrect reasoning, but only in reward-trained models, not instruction-tuned ones. We leverage this as a confidence signal to adaptively halt reasoning early, achieving 25-50% token reduction without sacrificing accuracy on math reasoning benchmarks.
arXiv Workshop

News

Sep 2026Two papers accepted as posters at NeurIPS 2026 — Frontier Coding Agents Use Metaprogramming (Main Track) and EsoLang-Bench (Evaluations and Datasets Track).
Aug 2026Started my MS CS at Georgia Tech and graduate research advised by Prof. Chao Zhang.
Apr 2026Admitted to Georgia Tech MS CS (Fall 2026)
Apr 2026EsoLang-Bench published at ICLR 2026 Workshops ICBINB & LLM Reasoning
Apr 2026LLM Deception in Auctions published at ICLR 2026 Workshop MALGAI
Dec 2025Two papers accepted at NeurIPS 2025 Workshops (FoRLM, Efficient Reasoning)
Oct 2025Ripple Effect Protocol published, part of Project Iceberg
Jun 2025Joined Lossfunk AI Lab as Researcher
Feb 2024Started research collaboration with MIT Media Lab
May 2023Joined Ikigai Labs as ML Engineer
Jul 2022MLH Fellowship - Hubble Protocol (Solana/Rust)
May 2022Google Summer of Code - NumFOCUS (Julia)

Get in Touch

Interested in collaborating, have research questions, or just want to chat? Drop me a message.