Privately Aligning Language Models with Reinforcement Learning

Positioned between pre-training and user deployment, aligning large language models (LLMs) through reinforcement learning (RL) has emerged as a prevailing strategy for training instruction following-models such as ChatGPT. In this work, we initiate the study of privacy-preserving alignment of LLMs through Differential Privacy (DP) in conjunction with RL. Following the influential work of Ziegler et al. (2020), we study two dominant paradigms: (i) alignment via RL without human in the loop (e.g., positive review generation) and (ii) alignment via RL from human feedback (RLHF) (e.g., summarization in a human-preferred way). We give a new DP framework to achieve alignment via RL, and prove its correctness. Our experimental results validate the effectiveness of our approach, offering competitive utility while ensuring strong privacy protections.

KITAB: Evaluating LLMs on Constraint Satisfaction for Information Retrieval

We study the ability of state-of-the art models to answer constraint satisfaction queries for information retrieval (e.g., 'a list of ice cream shops in San Diego'). In the past, such queries were considered to be tasks that could only be solved via web-search or knowledge bases. More recently, large language models (LLMs) have demonstrated initial emergent abilities in this task. However, many current retrieval benchmarks are either saturated or do not measure constraint satisfaction. Motivated by rising concerns around factual incorrectness and hallucinations of LLMs, we present KITAB, a new dataset for measuring constraint satisfaction abilities of language models. KITAB consists of book-related data across more than 600 authors and 13,000 queries, and also offers an associated dynamic data collection and constraint verification approach for acquiring similar test data for other authors. Our extended experiments on GPT4 and GPT3.5 characterize and decouple common failure modes across dimensions such as information popularity, constraint types, and context availability. Results show that in the absence of context, models exhibit severe limitations as measured by irrelevant information, factual errors, and incompleteness, many of which exacerbate as information popularity decreases. While context availability mitigates irrelevant information, it is not helpful for satisfying constraints, identifying fundamental barriers to constraint satisfaction. We open source our contributions to foster further research on improving constraint satisfaction abilities of future models.

Why Train More? Effective and Efficient Membership Inference via Memorization

Membership Inference Attacks (MIAs) aim to identify specific data samples within the private training dataset of machine learning models, leading to serious privacy violations and other sophisticated threats. Many practical black-box MIAs require query access to the data distribution (the same distribution where the private data is drawn) to train shadow models. By doing so, the adversary obtains models trained "with" or "without" samples drawn from the distribution, and analyzes the characteristics of the samples under consideration. The adversary is often required to train more than hundreds of shadow models to extract the signals needed for MIAs; this becomes the computational overhead of MIAs. In this paper, we propose that by strategically choosing the samples, MI adversaries can maximize their attack success while minimizing the number of shadow models. First, our motivational experiments suggest memorization as the key property explaining disparate sample vulnerability to MIAs. We formalize this through a theoretical bound that connects MI advantage with memorization. Second, we show sample complexity bounds that connect the number of shadow models needed for MIAs with memorization. Lastly, we confirm our theoretical arguments with comprehensive experiments; by utilizing samples with high memorization scores, the adversary can (a) significantly improve its efficacy regardless of the MIA used, and (b) reduce the number of shadow models by nearly two orders of magnitude compared to state-of-the-art approaches.

Diversity of Thought Improves Reasoning Abilities of Large Language Models

Large language models (LLMs) are documented to struggle in settings that require complex reasoning. Nevertheless, instructing the model to break down the problem into smaller reasoning steps (Wei et al., 2022), or ensembling various generations through modifying decoding steps (Wang et al., 2023) boosts performance. Current methods assume that the input prompt is fixed and expect the decoding strategies to introduce the diversity needed for ensembling. In this work, we relax this assumption and discuss how one can create and leverage variations of the input prompt as a means to diversity of thought to improve model performance. We propose a method that automatically improves prompt diversity by soliciting feedback from the LLM to ideate approaches that fit for the problem. We then ensemble the diverse prompts in our method DIV-SE (DIVerse reasoning path Self-Ensemble) across multiple inference calls. We also propose a cost-effective alternative where diverse prompts are used within a single inference call; we call this IDIV-SE (In-call DIVerse reasoning path Self-Ensemble). Under a fixed generation budget, DIV-SE and IDIV-SE outperform the previously discussed baselines using both GPT-3.5 and GPT-4 on several reasoning benchmarks, without modifying the decoding process. Additionally, DIV-SE advances state-of-the-art performance on recent planning benchmarks (Valmeekam et al., 2023), exceeding the highest previously reported accuracy by at least 29.6 percentage points on the most challenging 4/5 Blocksworld task. Our results shed light on how to enforce prompt diversity toward LLM reasoning and thereby improve the pareto frontier of the accuracy-cost trade-off.

Teaching Language Models to Hallucinate Less with Synthetic Tasks

Large language models (LLMs) frequently hallucinate on abstractive summarization tasks such as document-based question-answering, meeting summarization, and clinical report generation, even though all necessary information is included in context. However, optimizing LLMs to hallucinate less on these tasks is challenging, as hallucination is hard to efficiently evaluate at each optimization step. In this work, we show that reducing hallucination on a synthetic task can also reduce hallucination on real-world downstream tasks. Our method, SynTra, first designs a synthetic task where hallucinations are easy to elicit and measure. It next optimizes the LLM's system message via prefix-tuning on the synthetic task, and finally transfers the system message to realistic, hard-to-optimize tasks. Across three realistic abstractive summarization tasks, SynTra reduces hallucination for two 13B-parameter LLMs using only a synthetic retrieval task for supervision. We also find that optimizing the system message rather than the model weights can be critical; fine-tuning the entire model on the synthetic task can counterintuitively increase hallucination. Overall, SynTra demonstrates that the extra flexibility of working with synthetic data can help mitigate undesired behaviors in practice.

Attention Satisfies: A Constraint-Satisfaction Lens on Factual Errors of Language Models

We investigate the internal behavior of Transformer-based Large Language Models (LLMs) when they generate factually incorrect text. We propose modeling factual queries as Constraint Satisfaction Problems and use this framework to investigate how the model interacts internally with factual constraints. Specifically, we discover a strong positive relation between the model's attention to constraint tokens and the factual accuracy of its responses. In our curated suite of 11 datasets with over 40,000 prompts, we study the task of predicting factual errors with the Llama-2 family across all scales (7B, 13B, 70B). We propose SAT Probe, a method probing self-attention patterns, that can predict constraint satisfaction and factual errors, and allows early error identification. The approach and findings demonstrate how using the mechanistic understanding of factuality in LLMs can enhance reliability.

Sparks of Artificial General Intelligence: Early experiments with GPT-4

Artificial intelligence (AI) researchers have been developing and refining large language models (LLMs) that exhibit remarkable capabilities across a variety of domains and tasks, challenging our understanding of learning and cognition. The latest model developed by OpenAI, GPT-4, was trained using an unprecedented scale of compute and data. In this paper, we report on our investigation of an early version of GPT-4, when it was still in active development by OpenAI. We contend that (this early version of) GPT-4 is part of a new cohort of LLMs (along with ChatGPT and Google's PaLM for example) that exhibit more general intelligence than previous AI models. We discuss the rising capabilities and implications of these models. We demonstrate that, beyond its mastery of language, GPT-4 can solve novel and difficult tasks that span mathematics, coding, vision, medicine, law, psychology and more, without needing any special prompting. Moreover, in all of these tasks, GPT-4's performance is strikingly close to human-level performance, and often vastly surpasses prior models such as ChatGPT. Given the breadth and depth of GPT-4's capabilities, we believe that it could reasonably be viewed as an early (yet still incomplete) version of an artificial general intelligence (AGI) system. In our exploration of GPT-4, we put special emphasis on discovering its limitations, and we discuss the challenges ahead for advancing towards deeper and more comprehensive versions of AGI, including the possible need for pursuing a new paradigm that moves beyond next-word prediction. We conclude with reflections on societal influences of the recent technological leap and future research directions.

Verifiable and Provably Secure Machine Unlearning

Machine unlearning aims to remove points from the training dataset of a machine learning model after training; for example when a user requests their data to be deleted. While many machine unlearning methods have been proposed, none of them enable users to audit the unlearning procedure and verify that their data was indeed unlearned. To address this, we define the first cryptographic framework to formally capture the security of verifiable machine unlearning. While our framework is generally applicable to different approaches, its advantages are perhaps best illustrated by our instantiation for the canonical approach to unlearning: retraining the model without the data to be unlearned. In our cryptographic protocol, the server first computes a proof that the model was trained on a dataset~$D$. Given a user data point $d$, the server then computes a proof of unlearning that shows that $d \notin D$. We realize our protocol using a SNARK and Merkle trees to obtain proofs of update and unlearning on the data. Based on cryptographic assumptions, we then present a formal game-based proof that our instantiation is secure. Finally, we validate the practicality of our constructions for unlearning in linear regression, logistic regression, and neural networks.

On the Fundamental Limits of Formally Proving Robustness in Proof-of-Learning

Proof-of-learning (PoL) proposes a model owner use machine learning training checkpoints to establish a proof of having expended the necessary compute for training. The authors of PoL forego cryptographic approaches and trade rigorous security guarantees for scalability to deep learning by being applicable to stochastic gradient descent and adaptive variants. This lack of formal analysis leaves the possibility that an attacker may be able to spoof a proof for a model they did not train. We contribute a formal analysis of why the PoL protocol cannot be formally (dis)proven to be robust against spoofing adversaries. To do so, we disentangle the two roles of proof verification in PoL: (a) efficiently determining if a proof is a valid gradient descent trajectory, and (b) establishing precedence by making it more expensive to craft a proof after training completes (i.e., spoofing). We show that efficient verification results in a tradeoff between accepting legitimate proofs and rejecting invalid proofs because deep learning necessarily involves noise. Without a precise analytical model for how this noise affects training, we cannot formally guarantee if a PoL verification algorithm is robust. Then, we demonstrate that establishing precedence robustly also reduces to an open problem in learning theory: spoofing a PoL post hoc training is akin to finding different trajectories with the same endpoint in non-convex learning. Yet, we do not rigorously know if priori knowledge of the final model weights helps discover such trajectories. We conclude that, until the aforementioned open problems are addressed, relying more heavily on cryptography is likely needed to formulate a new class of PoL protocols with formal robustness guarantees. In particular, this will help with establishing precedence. As a by-product of insights from our analysis, we also demonstrate two novel attacks against PoL.

Generative Extraction of Audio Classifiers for Speaker Identification

It is perhaps no longer surprising that machine learning models, especially deep neural networks, are particularly vulnerable to attacks. One such vulnerability that has been well studied is model extraction: a phenomenon in which the attacker attempts to steal a victim's model by training a surrogate model to mimic the decision boundaries of the victim model. Previous works have demonstrated the effectiveness of such an attack and its devastating consequences, but much of this work has been done primarily for image and text processing tasks. Our work is the first attempt to perform model extraction on {\em audio classification models}. We are motivated by an attacker whose goal is to mimic the behavior of the victim's model trained to identify a speaker. This is particularly problematic in security-sensitive domains such as biometric authentication. We find that prior model extraction techniques, where the attacker \textit{naively} uses a proxy dataset to attack a potential victim's model, fail. We therefore propose the use of a generative model to create a sufficiently large and diverse pool of synthetic attack queries. We find that our approach is able to extract a victim's model trained on \texttt{LibriSpeech} using queries synthesized with a proxy dataset based off of \texttt{VoxCeleb}; we achieve a test accuracy of 84.41\% with a budget of 3 million queries.