We introduce 'FactCheck Editor', an advanced text editor designed to automate fact-checking and correct factual inaccuracies. Given the widespread issue of misinformation, often a result of unintentional mistakes by content creators, our tool aims to address this challenge. It supports over 90 languages and utilizes transformer models to assist humans in the labor-intensive process of fact verification. This demonstration showcases a complete workflow that detects text claims in need of verification, generates relevant search engine queries, and retrieves appropriate documents from the web. It employs Natural Language Inference (NLI) to predict the veracity of claims and uses LLMs to summarize the evidence and suggest textual revisions to correct any errors in the text. Additionally, the effectiveness of models used in claim detection and veracity assessment is evaluated across multiple languages.

While Reinforcement Learning (RL) has been proven essential for tuning large language models (LLMs), it can lead to reward over-optimization (ROO). Existing approaches address ROO by adding KL regularization, requiring computationally expensive hyperparameter tuning. Additionally, KL regularization focuses solely on regularizing the language policy, neglecting a potential source of regularization: the reward function itself. Inspired by demonstration-guided RL, we here introduce the Reward Calibration from Demonstration (RCfD), which leverages human demonstrations and a reward model to recalibrate the reward objective. Formally, given a prompt, the RCfD objective minimizes the distance between the demonstrations' and LLM's rewards rather than directly maximizing the reward function. This objective shift avoids incentivizing the LLM to exploit the reward model and promotes more natural and diverse language generation. We show the effectiveness of RCfD on three language tasks, which achieves comparable performance to carefully tuned baselines while mitigating ROO.

We extend the context length of Llama-3-8B-Instruct from 8K to 80K via QLoRA fine-tuning. The entire training cycle is super efficient, which takes 8 hours on one 8xA800 (80G) GPU machine. The resulted model exhibits superior performances across a broad range of evaluation tasks, such as NIHS, topic retrieval, and long-context language understanding; meanwhile, it also well preserves the original capability over short contexts. The dramatic context extension is mainly attributed to merely 3.5K synthetic training samples generated by GPT-4 , which indicates the LLMs' inherent (yet largely underestimated) potential to extend its original context length. In fact, the context length could be extended far beyond 80K with more computation resources. Therefore, the team will publicly release the entire resources (including data, model, data generation pipeline, training code) so as to facilitate the future research from the community: \url{https://github.com/FlagOpen/FlagEmbedding}.

Many of the recent capabilities demonstrated by Large Language Models (LLMs) arise primarily from their ability to exploit contextual information. In this paper, we explore ways to improve reasoning capabilities of LLMs through (1) exploration of different chains of thought and (2) validation of the individual steps of the reasoning process. We propose three general principles that a model should adhere to while reasoning: (i) Relevance, (ii) Mathematical Accuracy, and (iii) Logical Consistency. We apply these constraints to the reasoning steps generated by the LLM to improve the accuracy of the final generation. The constraints are applied in the form of verifiers: the model itself is asked to verify if the generated steps satisfy each constraint. To further steer the generations towards high-quality solutions, we use the perplexity of the reasoning steps as an additional verifier. We evaluate our method on 4 distinct types of reasoning tasks, spanning a total of 9 different datasets. Experiments show that our method is always better than vanilla generation, and, in 6 out of the 9 datasets, it is better than best-of N sampling which samples N reasoning chains and picks the lowest perplexity generation.

Large language models such as GPT and Llama are trained with a next-token prediction loss. In this work, we suggest that training language models to predict multiple future tokens at once results in higher sample efficiency. More specifically, at each position in the training corpus, we ask the model to predict the following n tokens using n independent output heads, operating on top of a shared model trunk. Considering multi-token prediction as an auxiliary training task, we measure improved downstream capabilities with no overhead in training time for both code and natural language models. The method is increasingly useful for larger model sizes, and keeps its appeal when training for multiple epochs. Gains are especially pronounced on generative benchmarks like coding, where our models consistently outperform strong baselines by several percentage points. Our 13B parameter models solves 12 % more problems on HumanEval and 17 % more on MBPP than comparable next-token models. Experiments on small algorithmic tasks demonstrate that multi-token prediction is favorable for the development of induction heads and algorithmic reasoning capabilities. As an additional benefit, models trained with 4-token prediction are up to 3 times faster at inference, even with large batch sizes.

The standard way to study Large Language Models (LLMs) with benchmarks or psychology questionnaires is to provide many different queries from similar minimal contexts (e.g. multiple choice questions). However, due to LLMs' highly context-dependent nature, conclusions from such minimal-context evaluations may be little informative about the model's behavior in deployment (where it will be exposed to many new contexts). We argue that context-dependence (specifically, value stability) should be studied a specific property of LLMs and used as another dimension of LLM comparison (alongside others such as cognitive abilities, knowledge, or model size). We present a case-study on the stability of value expression over different contexts (simulated conversations on different topics) as measured using a standard psychology questionnaire (PVQ) and on behavioral downstream tasks. Reusing methods from psychology, we study Rank-order stability on the population (interpersonal) level, and Ipsative stability on the individual (intrapersonal) level. We consider two settings (with and without instructing LLMs to simulate particular personas), two simulated populations, and three downstream tasks. We observe consistent trends in the stability of models and model families - Mixtral, Mistral, GPT-3.5 and Qwen families are more stable than LLaMa-2 and Phi. The consistency of these trends implies that some models exhibit higher value-stability than others, and that value stability can be estimated with the set of introduced methodological tools. When instructed to simulate particular personas, LLMs exhibit low Rank-Order stability, which further diminishes with conversation length. This highlights the need for future research on LLMs that coherently simulate different personas. This paper provides a foundational step in that direction, and, to our knowledge, it is the first study of value stability in LLMs.

The Financial Times has become the latest news organization to strike a deal with OpenAI. In a joint announcement on Monday, the Financial Times and OpenAI said that maker of ChatGPT will use the Financial Times' journalism to train its AI models and collaborate on developing new AI products and features for the publication's readers. "It is right, of course, that AI platforms pay publishers for the use of their material," said Financial Times CEO John Ridding in a statement and added that the Times is "committed to human journalism." Neither company disclosed the financial terms of the agreement. Earlier this year, The Information reported that OpenAI offers publishers between 1 million and 5 million a year to license their content to train its AI models.

The Financial Times has struck a deal with ChatGPT developer OpenAI that allows its content to be used in training artificial intelligence systems. The FT will receive an undisclosed payment as part of the deal, which is the latest to be agreed between OpenAI and news publishers. Under the arrangement, ChatGPT users will receive summaries and quotes from FT journalism, as well as links to articles, in responses to prompts, where appropriate. John Ridding, the chief executive of the FT Group, said it was "right" that AI companies paid publishers for their material. The New York Times is suing OpenAI and its largest investor, Microsoft, over use of its content to train large language models, the technology that underpins chatbots like ChatGPT.

Throughout a conversation, the way participants interact with each other is in constant flux: their tones may change, they may resort to different strategies to convey their points, or they might alter their interaction patterns. An understanding of these dynamics can complement that of the actual facts and opinions discussed, offering a more holistic view of the trajectory of the conversation: how it arrived at its current state and where it is likely heading. In this work, we introduce the task of summarizing the dynamics of conversations, by constructing a dataset of human-written summaries, and exploring several automated baselines. We evaluate whether such summaries can capture the trajectory of conversations via an established downstream task: forecasting whether an ongoing conversation will eventually derail into toxic behavior. We show that they help both humans and automated systems with this forecasting task. Humans make predictions three times faster, and with greater confidence, when reading the summaries than when reading the transcripts. Furthermore, automated forecasting systems are more accurate when constructing, and then predicting based on, summaries of conversation dynamics, compared to directly predicting on the transcripts.

Despite widespread applications of knowledge graphs (KGs) in various tasks such as question answering and intelligent conversational systems, existing KGs face two major challenges: information granularity and deficiency in timeliness. These hinder considerably the retrieval and analysis of in-context, fine-grained, and up-to-date knowledge from KGs, particularly in highly specialized themes (e.g., specialized scientific research) and rapidly evolving contexts (e.g., breaking news or disaster tracking). To tackle such challenges, we propose a theme-specific knowledge graph (i.e., ThemeKG), a KG constructed from a theme-specific corpus, and design an unsupervised framework for ThemeKG construction (named TKGCon). The framework takes raw theme-specific corpus and generates a high-quality KG that includes salient entities and relations under the theme. Specifically, we start with an entity ontology of the theme from Wikipedia, based on which we then generate candidate relations by Large Language Models (LLMs) to construct a relation ontology. To parse the documents from the theme corpus, we first map the extracted entity pairs to the ontology and retrieve the candidate relations. Finally, we incorporate the context and ontology to consolidate the relations for entity pairs. We observe that directly prompting GPT-4 for theme-specific KG leads to inaccurate entities (such as "two main types" as one entity in the query result) and unclear (such as "is", "has") or wrong relations (such as "have due to", "to start"). In contrast, by constructing the theme-specific KG step by step, our model outperforms GPT-4 and could consistently identify accurate entities and relations. Experimental results also show that our framework excels in evaluations compared with various KG construction baselines.