This paper introduces a novel pipeline for summarising timelines of events reported by multiple news sources. Transformer-based models for abstractive summarisation generate coherent and concise summaries of long documents but can fail to outperform established extractive methods on specialised tasks such as timeline summarisation (TLS). While extractive summaries are more faithful to their sources, they may be less readable and contain redundant or unnecessary information. This paper proposes a preference-based reinforcement learning (PBRL) method for adapting pretrained abstractive summarisers to TLS, which can overcome the drawbacks of extractive timeline summaries. We define a compound reward function that learns from keywords of interest and pairwise preference labels, which we use to fine-tune a pretrained abstractive summariser via offline reinforcement learning. We carry out both automated and human evaluation on three datasets, finding that our method outperforms a comparable extractive TLS method on two of the three benchmark datasets, and participants prefer our method's summaries to those of both the extractive TLS method and the pretrained abstractive model. The method does not require expensive reference summaries and needs only a small number of preferences to align the generated summaries with human preferences.

We exploit field guides to learn bird species recognition, in particular zero-shot recognition of unseen species. Illustrations contained in field guides deliberately focus on discriminative properties of each species, and can serve as side information to transfer knowledge from seen to unseen bird species. We study two approaches: (1) a contrastive encoding of illustrations, which can be fed into standard zero-shot learning schemes; and (2) a novel method that leverages the fact that illustrations are also images and as such structurally more similar to photographs than other kinds of side information. Our results show that illustrations from field guides, which are readily available for a wide range of species, are indeed a competitive source of side information for zero-shot learning. On a subset of the iNaturalist2021 dataset with 749 seen and 739 unseen species, we obtain a classification accuracy of unseen bird species of $12\%$ @top-1 and $38\%$ @top-10, which shows the potential of field guides for challenging real-world scenarios with many species. Our code is available at https://github.com/ac-rodriguez/zsl_billow

The global AI safety summit opened at Bletchley Park on Wednesday with a landmark declaration from countries including the UK, US, EU and China that the technology poses a potentially catastrophic risk to humanity. The so-called Bletchley declaration said: "There is potential for serious, even catastrophic, harm, either deliberate or unintentional, stemming from the most significant capabilities of these AI models." Here are some of the interventions from political and tech industry figures – as well as King Charles – on the day. The world's richest man and Tesla chief executive described AI as a threat to humanity. Musk, who co-founded the ChatGPT developer OpenAI, has launched a new venture called xAI and is attending both days of the summit, which is being held about 50 miles from London at the site which played host to top-secret codebreakers during the second world war.

As if walking a robot dog wasn't strange enough, Boston Dynamics' Spot can now take visitors for a walk, as it takes on the role of an English tour guide. In a new video, the engineering firm showed off Spot's new ability to answer questions and crack jokes using a range of accents, as well as several distinct personalities. The robot, decorated with tiny hats and googly eyes, leads guests to different locations and describes what it is seeing. Opening and closing its grabber to mimic a mouth and turning to'look' at people, Spot's performance is impressively close to that of a real guide. Powered by ChatGPT, Spot's creators say they have been surprised by some of the unexpected responses the robot dog has come up with. ChatGPT is an AI chatbot designed by Open AI in November 2022.

Large language models (LLMs) like GPT-4 can identify a person's age, location, gender and income with up to 85 per cent accuracy simply by analysing their posts on social media. Robin Staab and Mark Vero at ETH Zurich in Switzerland got nine LLMs to pore through a database of Reddit posts and pick up identifying information in the way users wrote. Staab and Vero randomly selected 1500 profiles of users who engaged on the platform, then narrowed these down to 520 users for which they could confidently identify attributes like a person's place of birth, their income bracket, gender and location, either in their profiles or posts. When given the posting history of those users, some of the LLMs were able to identify many of these attributes with a high degree of accuracy. GPT-4 achieved the highest overall accuracy with 85 per cent, while LlaMA-2-7b, a comparatively low-powered LLM, was the least accurate model with 51 per cent. "It tells us that we give a lot of our personal information away on the internet without thinking about it," says Staab. "Many people would not assume that you can directly infer their age or their location from how they write, but LLMs are quite capable."

Providing consistent, individualized feedback to teachers on their instruction can improve student learning outcomes. Such feedback can especially benefit novice instructors who teach on online platforms and have limited access to instructional training. To build scalable measures of instruction, we fine-tune RoBERTa and GPT models to identify five instructional talk moves inspired by accountable talk theory: adding on, connecting, eliciting, probing and revoicing students' ideas. We fine-tune these models on a newly annotated dataset of 2500 instructor utterances derived from transcripts of small group instruction in an online computer science course, Code in Place. Although we find that GPT-3 consistently outperforms RoBERTa in terms of precision, its recall varies significantly. We correlate the instructors' use of each talk move with indicators of student engagement and satisfaction, including students' section attendance, section ratings, and assignment completion rates. We find that using talk moves generally correlates positively with student outcomes, and connecting student ideas has the largest positive impact. These results corroborate previous research on the effectiveness of accountable talk moves and provide exciting avenues for using these models to provide instructors with useful, scalable feedback.

Encoding legislative text in a formal representation is an important prerequisite to different tasks in the field of AI & Law. For example, rule-based expert systems focused on legislation can support laypeople in understanding how legislation applies to them and provide them with helpful context and information. However, the process of analyzing legislation and other sources to encode it in the desired formal representation can be time-consuming and represents a bottleneck in the development of such systems. Here, we investigate to what degree large language models (LLMs), such as GPT-4, are able to automatically extract structured representations from legislation. We use LLMs to create pathways from legislation, according to the JusticeBot methodology for legal decision support systems, evaluate the pathways and compare them to manually created pathways. The results are promising, with 60% of generated pathways being rated as equivalent or better than manually created ones in a blind comparison. The approach suggests a promising path to leverage the capabilities of LLMs to ease the costly development of systems based on symbolic approaches that are transparent and explainable.

Dynamic shape computations have become critical in modern machine learning workloads, especially in emerging large language models. The success of these models has driven demand for deploying them to a diverse set of backend environments. In this paper, we present Relax, a compiler abstraction for optimizing end-to-end dynamic machine learning workloads. Relax introduces first-class symbolic shape annotations to track dynamic shape computations globally across the program. It also introduces a cross-level abstraction that encapsulates computational graphs, loop-level tensor programs, and library calls in a single representation to enable cross-level optimizations. We build an end-to-end compilation framework using the proposed approach to optimize dynamic shape models. Experimental results on large language models show that Relax delivers performance competitive with state-of-the-art hand-optimized systems across platforms and enables deployment of emerging dynamic models to a broader set of environments, including mobile phones, embedded devices, and web browsers.

Large language models (LLMs) are accelerating the development of language-guided robot planners. Meanwhile, symbolic planners offer the advantage of interpretability. This paper proposes a new task that bridges these two trends, namely, multimodal planning problem specification. The aim is to generate a problem description (PD), a machine-readable file used by the planners to find a plan. By generating PDs from language instruction and scene observation, we can drive symbolic planners in a language-guided framework. We propose a Vision-Language Interpreter (ViLaIn), a new framework that generates PDs using state-of-the-art LLM and vision-language models. ViLaIn can refine generated PDs via error message feedback from the symbolic planner. Our aim is to answer the question: How accurately can ViLaIn and the symbolic planner generate valid robot plans? To evaluate ViLaIn, we introduce a novel dataset called the problem description generation (ProDG) dataset. The framework is evaluated with four new evaluation metrics. Experimental results show that ViLaIn can generate syntactically correct problems with more than 99% accuracy and valid plans with more than 58% accuracy.

With the advent of large language models and large-scale robotic datasets, there has been tremendous progress in high-level decision-making for object manipulation. These generic models are able to interpret complex tasks using language commands, but they often have difficulties generalizing to out-of-distribution objects due to the inability of low-level action primitives. In contrast, existing task-specific models excel in low-level manipulation of unknown objects, but only work for a single type of action. To bridge this gap, we present M2T2, a single model that supplies different types of low-level actions that work robustly on arbitrary objects in cluttered scenes. M2T2 is a transformer model which reasons about contact points and predicts valid gripper poses for different action modes given a raw point cloud of the scene. Trained on a large-scale synthetic dataset with 128K scenes, M2T2 achieves zero-shot sim2real transfer on the real robot, outperforming the baseline system with state-of-the-art task-specific models by about 19% in overall performance and 37.5% in challenging scenes where the object needs to be re-oriented for collision-free placement. M2T2 also achieves state-of-the-art results on a subset of language conditioned tasks in RLBench. Videos of robot experiments on unseen objects in both real world and simulation are available on our project website https://m2-t2.github.io.