28 citations · 111 across the 25 of their papers we have counts for
26 papers
SpeechEditBench: A Bilingual Multi-Attribute Benchmark for Instruction-Guided Speech Editing
Hanlin Zhang, Daxin Tan, Dehua Tao +3
Instruction-guided speech editing requires a model to modify specified speech attributes while preserving non-target characteristics. Despite rapid progress in Speech Large Languag…
Peer-Predictive Self-Training for Language Model Reasoning
Shi Feng, Hanlin Zhang, Fan Nie +2
Mechanisms for continued self-improvement of language models without external supervision remain an open challenge. We propose Peer-Predictive Self-Training (PST), a label-free fin…
Next-token Preference: Scaling Unsupervised Reward Modeling
Jingxuan Fan, Yueying Li, Zhenting Qi +4
Learning from feedback advances the capabilities and safety of frontier models, but preference supervision remains costly and difficult to scale. We introduce next-token preference…
Prescriptive Scaling Reveals the Evolution of Language Model Capabilities
Hanlin Zhang, Jikai Jin, Vasilis Syrgkanis +1
Machine learning model performance improvements tend to arise from competition and application. For deployment, we consider prescriptive scaling laws: given a pre-training compute…
Weight Decay Improves Language Model Plasticity
Tessa Han, Sebastian Bordt, Hanlin Zhang +1
Large language models are typically trained in two broad phases: pretraining to produce a base model, followed by further training to improve downstream performance. However, hyper…
AlgoTune: Can Language Models Speed Up General-Purpose Numerical Programs?
Ori Press, Brandon Amos, Haoyu Zhao +21
Despite progress in language model (LM) capabilities, evaluations have thus far focused on models' performance on tasks that humans have previously solved, including in programming…
Discovering Hierarchical Latent Capabilities of Language Models via Causal Representation Learning
Jikai Jin, Vasilis Syrgkanis, Sham Kakade +1
Faithful evaluation of language model capabilities is crucial for deriving actionable insights that can inform model development. However, rigorous causal evaluations in this domai…
EvoLM: In Search of Lost Language Model Training Dynamics
Zhenting Qi, Fan Nie, Alexandre Alahi +6
Modern language model (LM) training has been divided into multiple stages, making it difficult for downstream developers to evaluate the impact of design choices made at each stage…
Connections between Schedule-Free Optimizers, AdEMAMix, and Accelerated SGD Variants
Depen Morwani, Nikhil Vyas, Hanlin Zhang +1
Recent advancements in deep learning optimization have introduced new algorithms, such as Schedule-Free optimizers, AdEMAMix, MARS and Lion which modify traditional momentum mechan…
How Does Critical Batch Size Scale in Pre-training?
Hanlin Zhang, Depen Morwani, Nikhil Vyas +5
Training large-scale models under given resources requires careful design of parallelism strategies. In particular, the efficiency notion of critical batch size (CBS), concerning t…
Eliminating Position Bias of Language Models: A Mechanistic Approach
Ziqi Wang, Hanlin Zhang, Xiner Li +6
Position bias has proven to be a prevalent issue of modern language models (LMs), where the models prioritize content based on its position within the given context. This bias ofte…
DataComp-LM: In search of the next generation of training sets for language models
Jeffrey Li, Alex Fang, Georgios Smyrnis +56
We introduce DataComp for Language Models (DCLM), a testbed for controlled dataset experiments with the goal of improving language models. As part of DCLM, we provide a standardize…
CoLoR-Filter: Conditional Loss Reduction Filtering for Targeted Language Model Pre-training
David Brandfonbrener, Hanlin Zhang, Andreas Kirsch +2
Selecting high-quality data for pre-training is crucial in shaping the downstream task performance of language models. A major challenge lies in identifying this optimal subset, a…
Follow My Instruction and Spill the Beans: Scalable Data Extraction from Retrieval-Augmented Generation Systems
Zhenting Qi, Hanlin Zhang, Eric Xing +2
Retrieval-Augmented Generation (RAG) improves pre-trained models by incorporating external knowledge at test time to enable customized adaptation. We study the risk of datastore le…
A Study on the Calibration of In-context Learning
Hanlin Zhang, Yi-Fan Zhang, Yaodong Yu +5
Accurate uncertainty quantification is crucial for the safe deployment of machine learning models, and prior research has demonstrated improvements in the calibration of modern lan…
Watermarks in the Sand: Impossibility of Strong Watermarking for Generative Models
Hanlin Zhang, Benjamin L. Edelman, Danilo Francati +3
Watermarking generative models consists of planting a statistical signal (watermark) in a model's output so that it can be later verified that the output was generated by the given…
Improved Logical Reasoning of Language Models via Differentiable Symbolic Programming
Hanlin Zhang, Jiani Huang, Ziyang Li +2
Pre-trained large language models (LMs) struggle to perform logical reasoning reliably despite advances in scale and compositionality. In this work, we tackle this challenge throug…
Do the Rewards Justify the Means? Measuring Trade-Offs Between Rewards and Ethical Behavior in the MACHIAVELLI Benchmark
Alexander Pan, Jun Shern Chan, Andy Zou +7
Artificial agents have traditionally been trained to maximize reward, which may incentivize power-seeking and deception, analogous to how next-token prediction in language models (…
Evaluating Step-by-Step Reasoning through Symbolic Verification
Yi-Fan Zhang, Hanlin Zhang, Li Erran Li +1
Pre-trained language models (LMs) have shown remarkable reasoning performance using explanations or chain-of-thoughts (CoT)) for in-context learning. On the other hand, these reaso…
A Closer Look at the Calibration of Differentially Private Learners
Hanlin Zhang, Xuechen Li, Prithviraj Sen +2
We systematically study the calibration of classifiers trained with differentially private stochastic gradient descent (DP-SGD) and observe miscalibration across a wide range of vi…
Exploring Transformer Backbones for Heterogeneous Treatment Effect Estimation
Yi-Fan Zhang, Hanlin Zhang, Zachary C. Lipton +2
Previous works on Treatment Effect Estimation (TEE) are not in widespread use because they are predominantly theoretical, where strong parametric assumptions are made but untractab…
Stochastic Neural Networks with Infinite Width are Deterministic
Liu Ziyin, Hanlin Zhang, Xiangming Meng +3
This work theoretically studies stochastic neural networks, a main type of neural network in use. We prove that as the width of an optimized stochastic neural network tends to infi…
Towards Principled Disentanglement for Domain Generalization
Hanlin Zhang, Yi-Fan Zhang, Weiyang Liu +3
A fundamental challenge for machine learning models is generalizing to out-of-distribution (OOD) data, in part due to spurious correlations. To tackle this challenge, we first form…
Toward Learning Human-aligned Cross-domain Robust Models by Countering Misaligned Features
Haohan Wang, Zeyi Huang, Hanlin Zhang +2
Machine learning has demonstrated remarkable prediction accuracy over i.i.d data, but the accuracy often drops when tested with data from another distribution. In this paper, we ai…