January 2026 - talks - lists.cs.princeton.edu

Shmuel Berman will present his General Exam "Beyond Long Context: Measuring Memory, Compression, and Calibration" on Friday, January 30, 2026 at 9:15 AM in CS 202 and via zoom.
by CS Grad Department 30 Jan '26

30 Jan '26

Shmuel Berman will present his General Exam "Beyond Long Context: Measuring Memory, Compression, and Calibration" on Friday, January 30, 2026 at 9:15 AM in CS 202 and via zoom. Zoom link: https://princeton.zoom.us/j/8227846962?omn=97734303297 Committee Members: Jia Deng (advisor), Zhuang Liu, Tom Griffiths Abstract: Humans reliably accumulate knowledge from experience over long time spans, but today’s LLMs and VLMs still struggle when the relevant evidence is distributed across long streams. Furthermore, even though humans frequently forget information, they are very cognizant of what they forget. We frame episodic memory as an online compression problem and investigate the memory capacity and compressive abilities of pretrained models and common architectures.. Our evaluation provides valuable information on how to improve the long-term memory of autonomous agents. Reading List: https://docs.google.com/document/d/1jsRPipttEXSFjk-tLyfAYpuZlztDuGEbaspba-x… Everyone is invited to attend the talk, and those faculty wishing to remain for the oral exam following are welcome to do so.

2 1

Berlin Chen will present his General Exam "Inference Efficiency for Sub-quadratic Models" on Friday, January 30, 2026 at 11:00 AM in Friend 108 and via zoom.
by CS Grad Department 29 Jan '26

29 Jan '26

Berlin Chen will present his General Exam "Inference Efficiency for Sub-quadratic Models" on Friday, January 30, 2026 at 11:00 AM in Friend 108 and via zoom. Zoom link: https://princeton.zoom.us/j/92265989878 Committee Members: Tri Dao (advisor), Elad Hazan, Kai Li Abstract: Recent progress in AI has witnessed a paradigm shift to test-time compute, where LLM inference commands an increasingly greater share of the compute budget. This shift presents a new opportunity for model design that tailors to the computational characteristic of inference. In this talk, I will present a recent work on improving the decoding efficiency of Mamba, which is a variant of subquadratic models based on State Space Models (SSMs). In particular, I will highlight a key challenge preventing subquadratic models from being hardware efficient during decoding. I will then propose an adjustment to the model that addresses the challenge, and demonstrate its key advantages from an inference-first perspective. I will further motivate the change by connecting it to classic SSMs. Drawing upon this connection, I will highlight two additional architectural adjustments to Mamba that are naturally motivated by classic SSM formulations and demonstrate their roles in improving model quality. Reading List: https://docs.google.com/document/d/1T2z9TI_StxTt9tMNawjSo5842qr8JSBP2zgl7oS… Everyone is invited to attend the talk, and those faculty wishing to remain for the oral exam following are welcome to do so.

2 1

Tim Kosfeld will present his General Exam "Computational modeling and integration of ATAC-seq" on Wednesday, January 28, 2026 at 12:00 PM in Bendheim House 103.
by CS Grad Department 23 Jan '26

23 Jan '26

Tim Kosfeld will present his General Exam "Computational modeling and integration of ATAC-seq" on Wednesday, January 28, 2026 at 12:00 PM in Bendheim House 103. Committee Members: Yuri Pritykin (advisor), Ben Raphael, Ellen Zhong Abstract: Assay for transposase-accessible chromatin with sequencing (ATAC-seq) has been widely used at single-cell (scATAC) and bulk levels to profile chromatin accessibility and explore the epigenomic properties in heterogeneous cell populations. Furthermore, transcription factors (TFs) bind to accessible sites, and scATAC-seq analysis could associate TF binding specificities, typically available as sequence motifs, with variance in chromatin accessibility. Here we present a new supervised TF motif regression method, MotifReg, for computational analysis and interpretation of scATAC-seq data. MotifReg can efficiently analyze large-scale scATAC-seq datasets and learn interpretable associations between TF motifs and accessible peaks. The resultant low-dimensional cell representations in the TF activity space facilitate downstream cell type-specific analyses of scATAC-seq data. The supervised formulation in MotifReg enables ranking TF motifs by significance of their association with chromatin accessibility, implicating cell type-specific TFs. Furthermore, by comparing MotifReg performance across data analysis settings, we develop guidelines and best practices for ATAC-seq peak calling, data preprocessing and normalization. We leverage these best practices to show that aggregation of biologically similar ATAC data allows for the discovery of chromosomal regions enriched for chromatin interaction, known motif binding, or ATAC co-accessibility. We provide a program to quickly and efficiently scrape, compile, and preprocess this data from online sources. These atlases improve per-nucleotide accessibility prediction in ATAC-seq through ChromBPNet and uniquely capture spacing preferences for known transcription factor complexes. Allowing a better understanding of transcription factor activity in accessible regions. Reading List: https://docs.google.com/document/d/1gnqGszJkHUIQisI3UL_mw-w_pCR2hD0dmL5N5OA… Everyone is invited to attend the talk, and those faculty wishing to remain for the oral exam following are welcome to do so.

1 0

Yebin Chon will present her FPO "LILAC: Enhancing Parallel Code Naturalness through Hybrid Compiler-LLM Translation" on Monday, January 26, 2026 at 1:30pm in CS 402.
by Gradinfo 20 Jan '26

20 Jan '26

Yebin Chon will present her FPO "LILAC: Enhancing Parallel Code Naturalness through Hybrid Compiler-LLM Translation" on Monday, January 26, 2026 at 1:30pm in CS 402. Readers: Margaret Martonosi, Zachary Kincaid, David August (adviser) Examiners: Sneha Goenka, Aarti Gupta, David August All are welcome to attend. Title: "LILAC: Enhancing Parallel Code Naturalness through Hybrid Compiler-LLM Translation" Abstract: Modern parallel hardware is becoming increasingly diverse, and there is a following increase in the heterogeneity of parallel programming models (PPMs), often tied to a particular parallel architecture. This brings a need for translation between different programming models to properly port and maintain code for various hardware systems. Source-to-source translators may automate this step and provide production compiler-level correctness guarantees. However, previous attempts to build translators that can go between different PPMs often result in unnatural code, hindering programmer comprehension and long-term maintenance. More recent Large Language Model (LLM) based methods produce highly natural, artful code, but lack the correctness guarantees of traditional compiler-based methods. One work provides a formal proof of correctness for the LLM-generated translation, but is severely limited in the code that it can process. This creates a critical trade-off: developers must choose between the correctness and applicability of compilers and the artfulness of LLMs, both of which are essential. This thesis provides a new way to bring the two methods together, creating cross-PPM translations that are both correct and natural. To achieve this, it introduces LILAC, a hybrid framework that combines the strengths of traceable and widely applicable compiler transformations and the stylistic capabilities of LLMs. LILAC builds on top of an existing compiler-based translator to provide functionally correct parallel translations and to simultaneously detect areas of potential unnaturalness. It then uses an LLM through an API to obtain suggestions for naturalizing the areas identified by the compiler. The compiler then checks these suggestions for correctness and applies the transformations to the code. In this way, LILAC performs several hybrid compiler-LLM passes to naturalize the code without compromising correctness. In a human study with 26 expert C/C++ programmers evaluating CUDA-to-OpenMP translations, LILAC-generated code was judged to be significantly more natural and maintainable than the output of the baseline compiler. The benchmarks used are unable to be translated by the formally verified LLM method. By building on a correct compiler translation, LILAC ensures that enhancements in naturalness do not come at the cost of correctness or applicability.

2 1

Yinghui He will present her General Exam "Skill-Targeted Adaptive Learning for LLMs" on Tuesday, January 20, 2026 at 3:00 PM in CS 401 and via zoom.
by CS Grad Department 15 Jan '26

15 Jan '26

Yinghui He will present her General Exam "Skill-Targeted Adaptive Learning for LLMs" on Tuesday, January 20, 2026 at 3:00 PM in CS 401 and via zoom. Zoom link: https://princeton.zoom.us/my/yinghuihe Committee Members: Sanjeev Arora (advisor), Danqi Chen, Zhuang Liu Abstract: Large language models (LLMs) can often “learn” at test time from carefully chosen in-context examples and at train time from supervised fine-tuning (SFT), yet both settings suffer from a common failure mode: generic instruction or data additions frequently yield diminishing returns and can even harm performance in smaller models. We propose skill-targeted adaptive learning as a unifying approach to closing these gaps by using an LLM teacher’s metacognition. First, we study adaptive skill-based in-context learning. We discovered that naive skill prompting can create a “cognitive overload” effect on easy questions by injecting unnecessary information. To address this, we present AdaptMI, which introduces skill-targeted in-context math instructions only when a model struggles, and AdaptMI+, which further targets examples to the specific skills missing from the model’s intermediate reasoning and final responses. Second, we explore skill-targeted adaptive training for settings where vanilla supervised fine-tuning saturates. We present STAT, a fine-tuning strategy in which a strong teacher labels each training instance with required skills, builds a Missing-Skill Profile by auditing the student’s errors, and then either (i) reweights/selects existing data or (ii) synthesizes new examples emphasizing missing skills. Together, these skill-targeted adaptive learning approaches provide a principled way to overcome performance saturation and systematically close capability gaps in language models. Reading List: https://docs.google.com/document/d/17ZEbYvjn-D7WVGy66YVyoRsUWGr4-pisie41dI4… Everyone is invited to attend the talk, and those faculty wishing to remain for the oral exam following are welcome to do so.

1 0

Yebin Chon will present her FPO "LILAC: Enhancing Parallel Code Naturalness through Hybrid Compiler-LLM Translation" on Monday, January 26, 2026 at 1:30pm in CS 402.
by Gradinfo 14 Jan '26

14 Jan '26

Yebin Chon will present her FPO "LILAC: Enhancing Parallel Code Naturalness through Hybrid Compiler-LLM Translation" on Monday, January 26 , 2026 at 1:30pm in CS 402. Readers: Margaret Martonosi, Zachary Kincaid, David August (adviser) Examiners: Sneha Goenka, Aarti Gupta, David August Title: "LILAC: Enhancing Parallel Code Naturalness through Hybrid Compiler-LLM Translation" All are welcome to attend. Abstract: Modern parallel hardware is becoming increasingly diverse, and there is a following increase in the heterogeneity of parallel programming models (PPMs), often tied to a particular parallel architecture. This brings a need for translation between different programming models in order to properly port and maintain code for various hardware systems. Source-to-source translators may automate this step and provide production compiler-level correctness guarantees. However, previous attempts to build translators that can go between different PPMs often result in unnatural and hard-to-read code, hindering programmer comprehension and long-term maintenance. More recent Large Language Model (LLM) based methods produce highly natural, artful code, but lack the correctness guarantees of traditional compiler-based methods. One work provides a formal proof of correctness for the LLM-generated translation, but is severely limited in the code that it can process. This creates a critical trade-off: developers must choose between the correctness and applicability of compilers and the artfulness of LLMs, both of which are essential. This thesis provides a new way to bring the two methods together, creating cross-PPM translations that are simultaneously correct and natural. To achieve this, it introduces LILAC, a hybrid framework that combines the strengths of traceable and widely applicable compiler transformations and the stylistic capabilities of LLMs. LILAC builds on top of an existing compiler-based translator to provide functionally correct parallel translations, and at the same time detect areas of potential unnaturalness. It then uses an LLM through an API to obtain suggestions for naturalizing the areas identified by the compiler. The compiler then checks these suggestions for correctness and applies the transformations to the code. In this way, LILAC performs several hybrid compiler-LLM passes to naturalize the code - refactoring control flow structures, improving variable names, and inserting comments - without compromising correctness. In a human study with 26 expert C/C++ programmers evaluating CUDA-to-OpenMP translations, LILAC-generated code was judged to be significantly more natural and maintainable than the output of the baseline compiler. The benchmarks used are unable to be translated by the formally verified LLM method. By building on a correct compiler translation, LILAC ensures that enhancements in naturalness do not come at the cost of correctness or applicability.

1 0

Chongyi Zheng will present his General Exam "Towards Unsupervised Pre-training for Reinforcement Learning" on Wednesday, January 14, 2026 at 11:00 AM in CS 301 and via zoom.
by CS Grad Department 09 Jan '26

09 Jan '26

Chongyi Zheng will present his General Exam "Towards Unsupervised Pre-training for Reinforcement Learning" on Wednesday, January 14, 2026 at 11:00 AM in CS 301 and via zoom. Zoom link: https://princeton.zoom.us/j/93974458862?pwd=1OO9PebfnbDu9vPe6XmmaP5TukAS7G.1 Committee Members: Benjamin Eysenbach (advisor), Peter Henderson, Karthik Narasimhan Abstract: Unsupervised reinforcement learning (RL) promises to solve new tasks without additional environment interaction, by leveraging prior experience and powerful representations. However, when and why unsupervised RL works remains poorly understood. In this presentation, I will explore one framework that enables unsupervised pre-training in RL, focusing on how temporal representations, training objectives, and policy adaptation jointly shape an agent’s ability to solve previously unseen tasks. I will propose new ways to understand a prior state-of-the-art unsupervised RL algorithm and develop a simplified version. I will also discuss some theoretical tools and diagnostic experiments that help us probe information acquired by unsupervised RL agents. Reading List: https://docs.google.com/document/d/1kEEdCLEyvFqUQ4bFDP-go0sBWNGIh7YBi2rwpSE… Everyone is invited to attend the talk, and those faculty wishing to remain for the oral exam following are welcome to do so.

1 0