arXiv subjects

Shushan Arakelyan

Publications and source records attributed to Shushan Arakelyan.

4 recordsLinked to original sources

arXiv2023

Exploring Distributional Shifts in Large Language Models for Code Analysis

We systematically study how three large language models with code capabilities - CodeT5, Codex, and ChatGPT - generalize to out-of-domain data. We consider two fundamental applications - code summarization, and code generation. We split data into domains following its natural boundaries - by an organization, by a project, and by a module within the software project. We establish that samples from each new domain present all the models with a significant challenge of distribution shift. We study how established methods adapt models to better generalize to new domains. Our experiments show that while multitask learning alone is a reasonable baseline, combining it with few-shot finetuning on examples retrieved from training data can achieve very strong performance. Moreover, this solution can outperform direct finetuning for very low-data scenarios. Finally, we consider variations of this approach to create a more broadly applicable method to adapt to multiple domains at once. We find that for code generation, a model adapted to multiple domains simultaneously performs on par with those adapted to a single domain

cs.CL↗

arXiv2022

NS3: Neuro-Symbolic Semantic Code Search

Semantic code search is the task of retrieving a code snippet given a textual description of its functionality. Recent work has been focused on using similarity metrics between neural embeddings of text and code. However, current language models are known to struggle with longer, compositional text, and multi-step reasoning. To overcome this limitation, we propose supplementing the query sentence with a layout of its semantic structure. The semantic layout is used to break down the final reasoning decision into a series of lower-level decisions. We use a Neural Module Network architecture to implement this idea. We compare our model - NS3 (Neuro-Symbolic Semantic Search) - to a number of baselines, including state-of-the-art semantic code retrieval methods, and evaluate on two datasets - CodeSearchNet and Code Search and Question Answering. We demonstrate that our approach results in more precise code retrieval, and we study the effectiveness of our modular design when handling compositional queries.

cs.LG↗

arXiv2021

Bin2vec: Learning Representations of Binary Executable Programs for Security Tasks

Tackling binary program analysis problems has traditionally implied manually defining rules and heuristics, a tedious and time-consuming task for human analysts. In order to improve automation and scalability, we propose an alternative direction based on distributed representations of binary programs with applicability to a number of downstream tasks. We introduce Bin2vec, a new approach leveraging Graph Convolutional Networks (GCN) along with computational program graphs in order to learn a high dimensional representation of binary executable programs. We demonstrate the versatility of this approach by using our representations to solve two semantically different binary analysis tasks - functional algorithm classification and vulnerability discovery. We compare the proposed approach to our own strong baseline as well as published results and demonstrate improvement over state-of-the-art methods for both tasks. We evaluated Bin2vec on 49191 binaries for the functional algorithm classification task, and on 30 different CWE-IDs including at least 100 CVE entries each for the vulnerability discovery task. We set a new state-of-the-art result by reducing the classification error by 40% compared to the source-code-based inst2vec approach, while working on binary code. For almost every vulnerability class in our dataset, our prediction accuracy is over 80% (and over 90% in multiple classes).

cs.CR↗

arXiv2018

Mining and Forecasting Career Trajectories of Music Artists

Many musicians, from up-and-comers to established artists, rely heavily on performing live to promote and disseminate their music. To advertise live shows, artists often use concert discovery platforms that make it easier for their fans to track tour dates. In this paper, we ask whether digital traces of live performances generated on those platforms can be used to understand career trajectories of artists. First, we present a new dataset we constructed by cross-referencing data from such platforms. We then demonstrate how this dataset can be used to mine and predict important career milestones for the musicians, such as signing by a major music label, or performing at a certain venue. Finally, we perform a temporal analysis of the bipartite artist-venue graph, and demonstrate that high centrality on this graph is correlated with success.

cs.SI↗