Search

We examine diverging preferences in human-labeled preference datasets. We develop a taxonomy of disagreement sources spanning 10 categories across four high-level classes -- task underspecification, response style, refusals, and annotation errors. We find that the majority of disagreements are in opposition with standard reward modeling approaches, which are designed with the assumption that annotator disagreement is noise. We then explore how these findings impact two areas of LLM development: reward modeling and evaluation. In our experiments, we demonstrate how standard reward modeling methods, like the Bradley-Terry model, fail to differentiate whether a given preference judgment is the result of unanimous agreement among annotators or the majority opinion among diverging user preferences. We also find that these tendencies are also echoed by popular LLM-as-Judge evaluation methods, which consistently identify a winning response in cases of diverging preferences. These findings highlight remaining challenges in LLM evaluations, which are greatly influenced by divisive features like response style, and in developing pluralistically aligned LLMs. To address these issues, we develop methods for identifying diverging preferences to mitigate their influence on evaluation and training.

Reward models are critical for aligning models to follow instructions, and are typically trained following one of two popular paradigms: Bradley-Terry style or Regression style. However, there is a lack of evidence that either approach is better than the other, when adequately matched for data. This is primarily because these approaches require data collected in different (but incompatible) formats, meaning that adequately matched data is not available in existing public datasets. To tackle this problem, we release preference annotations (designed for Bradley-Terry training) to complement existing ratings (designed for Regression style training) in the HelpSteer2 dataset. To improve data interpretability, preference annotations are accompanied with human-written justifications. Using this data, we conduct the first head-to-head comparison of Bradley-Terry and Regression models when adequately matched for data. Based on insights derived from such a comparison, we propose a novel approach to combine Bradley-Terry and Regression reward modeling. A Llama-3.1-70B-Instruct model tuned with this approach scores 94.1 on RewardBench, emerging top of more than 140 reward models as of 1 Oct 2024. We also demonstrate the effectiveness of this reward model at aligning models to follow instructions in RLHF. We open-source this dataset (CC-BY-4.0 license) at https://huggingface.co/datasets/nvidia/HelpSteer2 and openly release the trained Reward Model at https://huggingface.co/nvidia/Llama-3.1-Nemotron-70B-Reward, Comment: 26 pages, 3 figures

The development of LLMs has greatly enhanced the intelligence and fluency of question answering, while the emergence of retrieval enhancement has enabled models to better utilize external information. However, the presence of noise and errors in retrieved information poses challenges to the robustness of LLMs. In this work, to evaluate the model's performance under multiple interferences, we first construct a dataset based on machine reading comprehension datasets simulating various scenarios, including critical information absence, noise, and conflicts. To address the issue of model accuracy decline caused by noisy external information, we propose a data augmentation-based fine-tuning method to enhance LLM's robustness against noise. Additionally, contrastive learning approach is utilized to preserve the model's discrimination capability of external information. We have conducted experiments on both existing LLMs and our approach, the results are evaluated by GPT-4, which indicates that our proposed methods improve model robustness while strengthening the model's discrimination capability., Comment: This paper has been accepted by NLPCC-2024

The impressive capabilities of recent language models can be largely attributed to the multi-trillion token pretraining datasets that they are trained on. However, model developers fail to disclose their construction methodology which has lead to a lack of open information on how to develop effective pretraining sets. To address this issue, we perform the first systematic study across the entire pipeline of pretraining set construction. First, we run ablations on existing techniques for pretraining set development to identify which methods translate to the largest gains in model accuracy on downstream evaluations. Then, we categorize the most widely used data source, web crawl snapshots, across the attributes of toxicity, quality, type of speech, and domain. Finally, we show how such attribute information can be used to further refine and improve the quality of a pretraining set. These findings constitute an actionable set of steps that practitioners can use to develop high quality pretraining sets., Comment: Accepted as an oral presentation at EMNLP 2024

We release the Nemotron-4 340B model family, including Nemotron-4-340B-Base, Nemotron-4-340B-Instruct, and Nemotron-4-340B-Reward. Our models are open access under the NVIDIA Open Model License Agreement, a permissive model license that allows distribution, modification, and use of the models and its outputs. These models perform competitively to open access models on a wide range of evaluation benchmarks, and were sized to fit on a single DGX H100 with 8 GPUs when deployed in FP8 precision. We believe that the community can benefit from these models in various research studies and commercial applications, especially for generating synthetic data to train smaller language models. Notably, over 98% of data used in our model alignment process is synthetically generated, showcasing the effectiveness of these models in generating synthetic data. To further support open research and facilitate model development, we are also open-sourcing the synthetic data generation pipeline used in our model alignment process.

High-quality preference datasets are essential for training reward models that can effectively guide large language models (LLMs) in generating high-quality responses aligned with human preferences. As LLMs become stronger and better aligned, permissively licensed preference datasets, such as Open Assistant, HH-RLHF, and HelpSteer need to be updated to remain effective for reward modeling. Methods that distil preference data from proprietary LLMs such as GPT-4 have restrictions on commercial usage imposed by model providers. To improve upon both generated responses and attribute labeling quality, we release HelpSteer2, a permissively licensed preference dataset (CC-BY-4.0). Using a powerful internal base model trained on HelpSteer2, we are able to achieve the SOTA score (92.0%) on Reward-Bench's primary dataset, outperforming currently listed open and proprietary models, as of June 12th, 2024. Notably, HelpSteer2 consists of only ten thousand response pairs, an order of magnitude fewer than existing preference datasets (e.g., HH-RLHF), which makes it highly efficient for training reward models. Our extensive experiments demonstrate that reward models trained with HelpSteer2 are effective in aligning LLMs. In particular, we propose SteerLM 2.0, a model alignment approach that can effectively make use of the rich multi-attribute score predicted by our reward models. HelpSteer2 is available at https://huggingface.co/datasets/nvidia/HelpSteer2 and code is available at https://github.com/NVIDIA/NeMo-Aligner

AI-generated text detection has attracted increasing attention as powerful language models approach human-level generation. Limited work is devoted to detecting (partially) AI-paraphrased texts. However, AI paraphrasing is commonly employed in various application scenarios for text refinement and diversity. To this end, we propose a novel detection framework, paraphrased text span detection (PTD), aiming to identify paraphrased text spans within a text. Different from text-level detection, PTD takes in the full text and assigns each of the sentences with a score indicating the paraphrasing degree. We construct a dedicated dataset, PASTED, for paraphrased text span detection. Both in-distribution and out-of-distribution results demonstrate the effectiveness of PTD models in identifying AI-paraphrased text spans. Statistical and model analysis explains the crucial role of the surrounding context of the paraphrased text spans. Extensive experiments show that PTD models can generalize to versatile paraphrasing prompts and multiple paraphrased text spans. We release our resources at https://github.com/Linzwcs/PASTED., Comment: ACL 2024 Findings

Aligning Large Language Models (LLMs) with human values and preferences is essential for making them helpful and safe. However, building efficient tools to perform alignment can be challenging, especially for the largest and most competent LLMs which often contain tens or hundreds of billions of parameters. We create NeMo-Aligner, a toolkit for model alignment that can efficiently scale to a thousand GPUs for training the largest open-source LLMs such as Nemotron 4 340B and Llama 3.1 405B. NeMo-Aligner comes with highly optimized and scalable implementations for major paradigms of model alignment such as: Reinforcement Learning from Human Feedback (RLHF), Direct Preference Optimization (DPO), SteerLM, and Self-Play Fine-Tuning (SPIN). Additionally, our toolkit supports running most of the alignment techniques in a Parameter Efficient Fine-Tuning (PEFT) setting. NeMo-Aligner is designed for extensibility, allowing support for other alignment techniques with minimal effort. It is open-sourced with Apache 2.0 License and we invite community contributions at https://github.com/NVIDIA/NeMo-Aligner, Comment: 16 pages, 4 figures, Accepted to COLM 2024

Existing open-source helpfulness preference datasets do not specify what makes some responses more helpful and others less so. Models trained on these datasets can incidentally learn to model dataset artifacts (e.g. preferring longer but unhelpful responses only due to their length). To alleviate this problem, we collect HelpSteer, a multi-attribute helpfulness dataset annotated for the various aspects that make responses helpful. Specifically, our 37k-sample dataset has annotations for correctness, coherence, complexity, and verbosity in addition to overall helpfulness of responses. Training Llama 2 70B using the HelpSteer dataset with SteerLM technique produces a model that scores 7.54 on MT Bench, which is currently the highest score for open models that do not require training data from more powerful models (e.g. GPT4). We release this dataset with CC-BY-4.0 license at https://huggingface.co/datasets/nvidia/HelpSteer

Is it secure to measure the reliability of local models by similarity in federated learning (FL)? This paper delves into an unexplored security threat concerning applying similarity metrics, such as the L_2 norm, Euclidean distance, and cosine similarity, in protecting FL. We first uncover the deficiencies of similarity metrics that high-dimensional local models, including benign and poisoned models, may be evaluated to have the same similarity while being significantly different in the parameter values. We then leverage this finding to devise a novel untargeted model poisoning attack, Faker, which launches the attack by simultaneously maximizing the evaluated similarity of the poisoned local model and the difference in the parameter values. Experimental results based on seven datasets and eight defenses show that Faker outperforms the state-of-the-art benchmark attacks by 1.1-9.0X in reducing accuracy and 1.2-8.0X in saving time cost, which even holds for the case of a single malicious client with limited knowledge about the FL system. Moreover, Faker can degrade the performance of the global model by attacking only once. We also preliminarily explore extending Faker to other attacks, such as backdoor attacks and Sybil attacks. Lastly, we provide a model evaluation strategy, called the similarity of partial parameters (SPP), to defend against Faker. Given that numerous mechanisms in FL utilize similarity metrics to assess local models, this work suggests that we should be vigilant regarding the potential risks of using these metrics.

Just as computational simulations of atoms, molecules and cells have shaped the way we study the sciences, true-to-life simulations of human-like agents can be valuable tools for studying human behavior. We propose Humanoid Agents, a system that guides Generative Agents to behave more like humans by introducing three elements of System 1 processing: Basic needs (e.g. hunger, health and energy), Emotion and Closeness in Relationships. Humanoid Agents are able to use these dynamic elements to adapt their daily activities and conversations with other agents, as supported with empirical experiments. Our system is designed to be extensible to various settings, three of which we demonstrate, as well as to other elements influencing human behavior (e.g. empathy, moral values and cultural background). Our platform also includes a Unity WebGL game interface for visualization and an interactive analytics dashboard to show agent statuses over time. Our platform is available on https://www.humanoidagents.com/ and code is on https://github.com/HumanoidAgents/HumanoidAgents, Comment: Accepted at EMNLP System Demonstrations 2023

Model alignment with human preferences is an essential step in making Large Language Models (LLMs) helpful and consistent with human values. It typically consists of supervised fine-tuning (SFT) and reinforcement learning from human feedback (RLHF) stages. However, RLHF faces inherent limitations stemming from a complex training setup and its tendency to align the model with implicit values that end users cannot control at run-time. Moreover, reward models in RLHF stage commonly rely on single-dimensional feedback as opposed to explicit, multifaceted signals that indicate attributes such as helpfulness, humor, and toxicity. To address these limitations, we propose SteerLM, a supervised fine-tuning method that empowers end-users to control responses during inference. SteerLM conditions responses to conform to an explicitly defined multi-dimensional set of attributes, thereby empowering a steerable AI capable of generating helpful and high-quality responses while maintaining customizability. Experiments show that SteerLM trained on open source datasets generates responses that are preferred by human and automatic evaluators to many state-of-the-art baselines trained with RLHF while being much easier to train. Try SteerLM at https://huggingface.co/nvidia/SteerLM-llama2-13B, Comment: Findings of EMNLP 2023

Facilitated by mobile edge computing, client-edge-cloud hierarchical federated learning (HFL) enables communication-efficient model training in a widespread area but also incurs additional security and privacy challenges from intermediate model aggregations and remains the single point of failure issue. To tackle these challenges, we propose a blockchain-based HFL (BHFL) system that operates a permissioned blockchain among edge servers for model aggregation without the need for a centralized cloud server. The employment of blockchain, however, introduces additional overhead. To enable a compact and efficient workflow, we design a novel lightweight consensus algorithm, named Proof of Federated Edge Learning (PoFEL), to recycle the energy consumed for local model training. Specifically, the leader node is selected by evaluating the intermediate FEL models from all edge servers instead of other energy-wasting but meaningless calculations. This design thus improves the system efficiency compared with traditional BHFL frameworks. To prevent model plagiarism and bribery voting during the consensus process, we propose Hash-based Commitment and Digital Signature (HCDS) and Bayesian Truth Serum-based Voting (BTSV) schemes. Finally, we devise an incentive mechanism to motivate continuous contributions from clients to the learning task. Experimental results demonstrate that our proposed BHFL system with the corresponding consensus protocol and incentive mechanism achieves effectiveness, low computational cost, and fairness.

Cloud-edge-device hierarchical federated learning (HFL) has been recently proposed to achieve communication-efficient and privacy-preserving distributed learning. However, there exist several critical challenges, such as the single point of failure and potential stragglers in both edge servers and local devices. To resolve these issues, we propose a decentralized and straggler-tolerant blockchain-based HFL (BHFL) framework. Specifically, a Raft-based consortium blockchain is deployed on edge servers to provide a distributed and trusted computing environment for global model aggregation in BHFL. To mitigate the influence of stragglers on learning, we propose a novel aggregation method, HieAvg, which utilizes the historical weights of stragglers to estimate the missing submissions. Furthermore, we optimize the overall latency of BHFL by jointly considering the constraints of global model convergence and blockchain consensus delay. Theoretical analysis and experimental evaluation show that our proposed BHFL based on HieAvg can converge in the presence of stragglers, which performs better than the traditional methods even when the loss function is non-convex and the data on local devices are non-independent and identically distributed (non-IID).

Large language models (LLMs) have achieved human-level text generation, emphasizing the need for effective AI-generated text detection to mitigate risks like the spread of fake news and plagiarism. Existing research has been constrained by evaluating detection methods on specific domains or particular language models. In practical scenarios, however, the detector faces texts from various domains or LLMs without knowing their sources. To this end, we build a comprehensive testbed by gathering texts from diverse human writings and texts generated by different LLMs. Empirical results show challenges in distinguishing machine-generated texts from human-authored ones across various scenarios, especially out-of-distribution. These challenges are due to the decreasing linguistic distinctions between the two sources. Despite challenges, the top-performing detector can identify 86.54% out-of-domain texts generated by a new LLM, indicating the feasibility for application scenarios. We release our resources at https://github.com/yafuly/MAGE., Comment: ACL 2024

The widely used Fact-based Visual Question Answering (FVQA) dataset contains visually-grounded questions that require information retrieval using common sense knowledge graphs to answer. It has been observed that the original dataset is highly imbalanced and concentrated on a small portion of its associated knowledge graph. We introduce FVQA 2.0 which contains adversarial variants of test questions to address this imbalance. We show that systems trained with the original FVQA train sets can be vulnerable to adversarial samples and we demonstrate an augmentation scheme to reduce this vulnerability without human annotations., Comment: Accepted to EACL 2023 Findings

Although Metaverse has recently been widely studied, its practical application still faces many challenges. One of the severe challenges is the lack of sufficient resources for computing and communication on local devices, resulting in the inability to access the Metaverse services. To address this issue, this paper proposes a practical blockchain-based mobile edge computing (MEC) platform for resource sharing and optimal utilization to complete the requested offloading tasks, given the heterogeneity of servers' available resources and that of users' task requests. To be specific, we first elaborate the design of our proposed system and then dive into the task allocation mechanism to assign offloading tasks to proper servers. To solve the multiple task allocation (MTA) problem in polynomial time, we devise a learning-based algorithm. Since the objective function and constraints of MTA are significantly affected by the servers uploading the tasks, we reformulate it as a reinforcement learning problem and calculate the rewards for each state and action considering the influences of servers. Finally, numerous experiments are conducted to demonstrate the effectiveness and efficiency of our proposed system and algorithms.

With the development of mobile edge computing (MEC) and blockchain-based federated learning (BCFL), a number of studies suggest deploying BCFL on edge servers. In this case, resource-limited edge servers need to serve both mobile devices for their offloading tasks and the BCFL system for model training and blockchain consensus in a cost-efficient manner without sacrificing the service quality to any side. To address this challenge, this paper proposes a resource allocation scheme for edge servers, aiming to provide the optimal services with the minimum cost. Specifically, we first analyze the energy consumed by the MEC and BCFL tasks, and then use the completion time of each task as the service quality constraint. Then, we model the resource allocation challenge into a multivariate, multi-constraint, and convex optimization problem. To solve the problem in a progressive manner, we design two algorithms based on the alternating direction method of multipliers (ADMM) in both the homogeneous and heterogeneous situations with equal and on-demand resource distribution strategies, respectively. The validity of our proposed algorithms is proved via rigorous theoretical analysis. Through extensive experiments, the convergence and efficiency of our proposed resource allocation schemes are evaluated. To the best of our knowledge, this is the first work to investigate the resource allocation dilemma of edge servers for BCFL in MEC.

Knowledge graph (KG) based Collaborative Filtering is an effective approach to personalizing recommendation systems for relatively static domains such as movies and books, by leveraging structured information from KG to enrich both item and user representations. Motivated by the use of Transformers for understanding rich text in content-based filtering recommender systems, we propose Content-aware KG-enhanced Meta-preference Networks as a way to enhance collaborative filtering recommendation based on both structured information from KG as well as unstructured content features based on Transformer-empowered content-based filtering. To achieve this, we employ a novel training scheme, Cross-System Contrastive Learning, to address the inconsistency of the two very different systems and propose a powerful collaborative filtering model and a variant of the well-known NRMS system within this modeling framework. We also contribute to public domain resources through the creation of a large-scale movie-knowledge-graph dataset and an extension of the already public Amazon-Book dataset through incorporation of text descriptions crawled from external sources. We present experimental results showing that enhancing collaborative filtering with Transformer-based features derived from content-based filtering outperforms strong baseline systems, improving the ability of knowledge-graph-based collaborative filtering systems to exploit item content information.

Blockchain-based federated learning (BCFL) has recently gained tremendous attention because of its advantages such as decentralization and privacy protection of raw data. However, there has been few research focusing on the allocation of resources for clients in BCFL. In the BCFL framework where the FL clients and the blockchain miners are the same devices, clients broadcast the trained model updates to the blockchain network and then perform mining to generate new blocks. Since each client has a limited amount of computing resources, the problem of allocating computing resources into training and mining needs to be carefully addressed. In this paper, we design an incentive mechanism to assign each client appropriate rewards for training and mining, and then the client will determine the amount of computing power to allocate for each subtask based on these rewards using the two-stage Stackelberg game. After analyzing the utilities of the model owner (MO) (i.e., the BCFL task publisher) and clients, we transform the game model into two optimization problems, which are sequentially solved to derive the optimal strategies for both the MO and clients. Further, considering the fact that local training related information of each client may not be known by others, we extend the game model with analytical solutions to the incomplete information scenario. Extensive experimental results demonstrate the validity of our proposed schemes.

Advances in distributed machine learning can empower future communications and networking. The emergence of federated learning (FL) has provided an efficient framework for distributed machine learning, which, however, still faces many security challenges. Among them, model poisoning attacks have a significant impact on the security and performance of FL. Given that there have been many studies focusing on defending against model poisoning attacks, it is necessary to survey the existing work and provide insights to inspire future research. In this paper, we first classify defense mechanisms for model poisoning attacks into two categories: evaluation methods for local model updates and aggregation methods for the global model. Then, we analyze some of the existing defense strategies in detail. We also discuss some potential challenges and future research directions. To the best of our knowledge, we are the first to survey defense methods for model poisoning attacks in FL.

Mobile crowdsensing (MCS) counting on the mobility of massive workers helps the requestor accomplish various sensing tasks with more flexibility and lower cost. However, for the conventional MCS, the large consumption of communication resources for raw data transmission and high requirements on data storage and computing capability hinder potential requestors with limited resources from using MCS. To facilitate the widespread application of MCS, we propose a novel MCS learning framework leveraging on blockchain technology and the new concept of edge intelligence based on federated learning (FL), which involves four major entities, including requestors, blockchain, edge servers and mobile devices as workers. Even though there exist several studies on blockchain-based MCS and blockchain-based FL, they cannot solve the essential challenges of MCS with respect to accommodating resource-constrained requestors or deal with the privacy concerns brought by the involvement of requestors and workers in the learning process. To fill the gaps, four main procedures, i.e., task publication, data sensing and submission, learning to return final results, and payment settlement and allocation, are designed to address major challenges brought by both internal and external threats, such as malicious edge servers and dishonest requestors. Specifically, a mechanism design based data submission rule is proposed to guarantee the data privacy of mobile devices being truthfully preserved at edge servers; consortium blockchain based FL is elaborated to secure the distributed learning process; and a cooperation-enforcing control strategy is devised to elicit full payment from the requestor. Extensive simulations are carried out to evaluate the performance of our designed schemes.

With the technological advances in machine learning, effective ways are available to process the huge amount of data generated in real life. However, issues of privacy and scalability will constrain the development of machine learning. Federated learning (FL) can prevent privacy leakage by assigning training tasks to multiple clients, thus separating the central server from the local devices. However, FL still suffers from shortcomings such as single-point-failure and malicious data. The emergence of blockchain provides a secure and efficient solution for the deployment of FL. In this paper, we conduct a comprehensive survey of the literature on blockchained FL (BCFL). First, we investigate how blockchain can be applied to federal learning from the perspective of system composition. Then, we analyze the concrete functions of BCFL from the perspective of mechanism design and illustrate what problems blockchain addresses specifically for FL. We also survey the applications of BCFL in reality. Finally, we discuss some challenges and future research directions.

Personal attributes represent structured information about a person, such as their hobbies, pets, family, likes and dislikes. We introduce the tasks of extracting and inferring personal attributes from human-human dialogue, and analyze the linguistic demands of these tasks. To meet these challenges, we introduce a simple and extensible model that combines an autoregressive language model utilizing constrained attribute generation with a discriminative reranker. Our model outperforms strong baselines on extracting personal attributes as well as inferring personal attributes that are not contained verbatim in utterances and instead requires commonsense reasoning and lexical inferences, which occur frequently in everyday conversation. Finally, we demonstrate the benefit of incorporating personal attributes in social chit-chat and task-oriented dialogue settings., Comment: 12 pages, 3 figures

While many different aspects of human experiences have been studied by the NLP community, none has captured its full richness. We propose a new task to capture this richness based on an unlikely setting: movie characters. We sought to capture theme-level similarities between movie characters that were community-curated into 20,000 themes. By introducing a two-step approach that balances performance and efficiency, we managed to achieve 9-27\% improvement over recent paragraph-embedding based methods. Finally, we demonstrate how the thematic information learnt from movie characters can potentially be used to understand themes in the experience of people, as indicated on Reddit posts., Comment: To appear in Proceedings of the 2021 NAACL Workshop WNU: 3rd Workshop on Narrative Understanding

Online forums and social media platforms provide noisy but valuable data every day. In this paper, we propose a novel end-to-end neural network-based user embedding system, Author2Vec. The model incorporates sentence representations generated by BERT (Bidirectional Encoder Representations from Transformers) with a novel unsupervised pre-training objective, authorship classification, to produce better user embedding that encodes useful user-intrinsic properties. This user embedding system was pre-trained on post data of 10k Reddit users and was analyzed and evaluated on two user classification benchmarks: depression detection and personality classification, in which the model proved to outperform traditional count-based and prediction-based methods. We substantiate that Author2Vec successfully encoded useful user attributes and the generated user embedding performs well in downstream classification tasks without further finetuning.