@inproceedings{sharma2026schedbench,
title = {{SCHEDBench}: A Benchmark for Evaluating LLM Constraint Faithfulness in Natural-Language Combinatorial Scheduling},
author = {Sharma, Shrenil Shaun and Sharma, Avi},
booktitle = {{Findings of the Association for Computational Linguistics: EMNLP 2026}},
year = {2026},
url = {https://arxiv.org/abs/2608.00991}
}
×
|
Avi Sharma
Hi! I’m an undergrad studying Electrical Engineering and Computer Science (EECS) at the University of California, Berkeley and doing research at Berkeley Artificial Intelligence Research. I’m interested in robotics, reinforcement learning, and neuro-symbolic AI. |
![]() |
|---|---|
Publications and Preprints |
|
EMNLP 2026 Findings
We introduce SCHEDBench, a benchmark for testing whether LLMs can remain faithful to scheduling constraints in natural-language combinatorial planning tasks.
2026
|
|
MathNLP Workshop @ EMNLP 2026
We introduce SDDL, a neuro-symbolic framework for combinatorial scheduling, and demonstrate how formal intermediate representations can improve LLM planning capabilities.
×
YEAR
|
|
CONFERENCE YEAR
Worst Paper Award
PLACEHOLDER one-to-two sentence summary of the core problem, method, and main result.
×
YEAR
|
|
|
equal contribution
|
|
