Research
Research
Research
Research
2,583+ open-access research outputs.
We study when a sound arithmetic theory $\mathcal S{\supseteq}S^1_2$ with polynomial-time decidable axioms efficiently proves the bounded consistency statements $Con_{\mathcal S{+}\phi}(n)$ for a true…
To enhance LLMs' impact on math education, we need data on their mathematical prowess and biases across prompts. To fill this gap, we introduce MEDS (Math Education Digital Shadows) as a dataset mappi…
Democratic discourse analysis systems increasingly rely on multi-agent LLM pipelines in which distinct evaluator models are assigned adversarial roles to generate structured, multi-perspective assessm…
Large language models (LLMs) acquire most of their factual knowledge during the pre-training stage, through next token prediction. Subsequent stages of post-training often introduce new facts outwith …
To overcome the well-known memory bottleneck of AI chips, 3D stacked architectures that employ advanced packaging technology with high-density through-silicon vias (TSVs) pins have proven to be a prom…
Every RLHF-trained language model is shaped by a reward model, yet the mechanistic interpretability toolkit -- logit lens, direct logit attribution, activation patching, sparse autoencoders -- was bui…
We investigate linguistic biases in LLM-based restaurant and product recommendations given prompts varying across Southern American English (AE), Indian English (IE), and Code-Switched Hindi-English d…
Typographic prompt injection exploits vision language models' (VLMs) ability to read text rendered in images, posing a growing threat as VLMs power autonomous agents. Prior work typically focus on max…
Post-Training Quantization (PTQ) compresses large language models to low bit-widths using a small calibration set, and its quality depends strongly on which samples are chosen. We identify a failure m…
Autoregressive transformers make confident errors, but activation monitoring can catch them only if the model preserves an internal signal that output confidence does not expose. This preservation is …
Automatic induction of high-quality dictionaries is essential for building lexical resources, yet low-resource languages and dialects pose several challenges: limited access to annotators, high degree…
Hypernetwork-based methods such as Doc-to-LoRA internalize a document into an LLM's weights in a single forward pass, but they fail systematically on conflicts: when the document contradicts pretraini…
We study the organization of channel-level importance in transformer feed-forward networks (FFNs). Using a Fisher-style loss proxy (LP) based on activation-gradient second moments, we show that loss s…
Large language models (LLMs) have demonstrated strong performance on a wide range of software engineering tasks, including code generation and analysis. However, most prior work relies on cloud-based …
Health misinformation remains one of the most pressing challenges on social media, particularly when cultural traditions intersect with scientific-sounding claims. These dynamics are not only global b…
The mass-loss process of red supergiant (RSG) and asymptotic giant branch (AGB) stars and its relation to variability is poorly constrained. We study two evolved stars, the Mira-type AGB star R Car an…
Medication reconciliation at clinical handoffs is a high-stakes, error-prone process. Large language models are increasingly proposed to assist with this task using FHIR-structured patient records, bu…
We present a systematic empirical study of transformer compression through over 40 experiments on GPT-2 (124M parameters) and Mistral 7B (7.24B parameters). Our analysis covers spectral compression, b…
LLM agents deployed in production operate under operator-defined behavioral policies (system-prompt instructions such as prohibitions on credential disclosure, data exfiltration, and unauthorized outp…
This paper narrows the performance gap between small, specialized models and significantly larger general-purpose models through domain adaptation via continual pre-training and merging. We address th…
Free open-access publishing with Google Scholar indexing.
Submission Guide →