Pearl Abyss, the game's developer, issued a lengthy apology on X and detailed its corrective actions. The developer behind the open-world RPG Crimson Desert has issued an official apology after players discovered several instances of AI-generated art in the game. Pearl Abyss posted on X that it released the game with some 2D visual props that were made with experimental AI generative tools and forgot to replace them before launch. We would like to address questions regarding the use of AI in Crimson Desert. During development, some 2D visual props were created as part of early-stage iteration using experimental AI generative tools.

Recent work has shown it is possible to construct adversarial examples that cause aligned language models to emit harmful strings or perform harmful behavior.Existing attacks work either in the white-box setting (with full access to the model weights), or through: the phenomenon that adversarial examples crafted on one model often remain effective on other models.We improve on prior work with a attack that leverages API access to a remote language model to construct adversarial examples that cause the model to emit harmful strings with (much) higher probability than with transfer-only attacks.We validate our attack on GPT-3.5 and OpenAI's safety classifier; we can cause GPT-3.5 to emit harmful strings that current transfer attacks fail at, and we can evade the OpenAI and Llama Guard safety classifiers with nearly 100% probability.

Equivariant machine learning is an approach for designing deep learning models that respect the symmetries of the problem, with the aim of reducing model complexity and improving generalization. In this paper, we focus on an extension of shift equivariance, which is the basis of convolution networks on images, to general graphs. Unlike images, graphs do not have a natural notion of domain translation. Therefore, we consider the graph functional shifts as the symmetry group: the unitary operators that commute with the graph shift operator. Notably, such symmetries operate in the signal space rather than directly in the spatial space.We remark that each linear filter layer of a standard spectral graph neural network (GNN) commutes with graph functional shifts, but the activation function breaks this symmetry. Instead, we propose nonlinear spectral filters (NLSFs) that are fully equivariant to graph functional shifts and show that they have universal approximation properties. The proposed NLSFs are based on a new form of spectral domain that is transferable between graphs. We demonstrate the superior performance of NLSFs over existing spectral GNNs in node and graph classification benchmarks.

With the rapid development of deep learning, the increasing complexity and scale of parameters make training a new model increasingly resource-intensive. In this paper, we start from the classic convolutional neural network (CNN) and explore a paradigm that does not require training to obtain new models. Similar to the birth of CNN inspired by receptive fields in the biological visual system, we draw inspiration from the information subsystem pathways in the biological visual system and propose Model Disassembling and Assembling (MDA). During model disassembling, we introduce the concept of relative contribution and propose a component locating technique to extract task-aware components from trained CNN classifiers. For model assembling, we present the alignment padding strategy and parameter scaling strategy to construct a new model tailored for a specific task, utilizing the disassembled task-aware components.The entire process is akin to playing with LEGO bricks, enabling arbitrary assembly of new models, and providing a novel perspective for model creation and reuse. Extensive experiments showcase that task-aware components disassembled from CNN classifiers or new models assembled using these components closely match or even surpass the performance of the baseline,demonstrating its promising results for model reuse. Furthermore, MDA exhibits diverse potential applications, with comprehensive experiments exploring model decision route analysis, model compression, knowledge distillation, and more.

Data-driven deep learning models are transforming global weather forecasting. It is an open question if this success can extend to climate modeling, where the complexity of the data and long inference rollouts pose significant challenges. Here, we present the first conditional generative model that produces accurate and physically consistent global climate ensemble simulations by emulating a coarse version of the United States' primary operational global forecast model, FV3GFS.Our model integrates the dynamics-informed diffusion framework (DYffusion) with the Spherical Fourier Neural Operator (SFNO) architecture, enabling stable 100-year simulations at 6-hourly timesteps while maintaining low computational overhead compared to single-step deterministic baselines.The model achieves near gold-standard performance for climate model emulation, outperforming existing approaches and demonstrating promising ensemble skill.This work represents a significant advance towards efficient, data-driven climate simulations that can enhance our understanding of the climate system and inform adaptation strategies.

Large Language Models (LLMs) have seen an impressive wave of advances, withmodels now excelling in a variety of tasks, such as mathematical reasoning andprogram synthesis. However, their potential to effectively use tools via API callsremains unfulfilled. This is a challenging task even for today's state-of-the-artLLMs such as GPT-4 largely due to their unawareness of what APIs are availableand how to use them in a frequently updated tool set. We develop Gorilla, afinetuned LLaMA model that surpasses the performance of GPT-4 on writing APIcalls. Trained with the novel Retriever Aware Training (RAT), when combinedwith a document retriever, Gorilla demonstrates a strong capability to adapt totest-time document changes, allowing flexible user updates or version changes.It also substantially mitigates the issue of hallucination, commonly encounteredwhen prompting LLMs directly. To evaluate the model's ability, we introduceAPIBench, a comprehensive dataset consisting of HuggingFace, TorchHub, andTensorHub APIs. The successful integration of the retrieval system with Gorillademonstrates the potential for LLMs to use tools more accurately, keep up withfrequently updated documentation, and consequently increase the reliability andapplicability of their outputs. Gorilla's code, model, data, and demo are availableat: https://gorilla.cs.berkeley.edu

Recent advances in large language models have shown capabilities that are extraordinary and near-superhuman. These models operate with such complexity that reliably evaluating and aligning them proves challenging for humans. This leads to the natural question: can guidance from weak models (like humans) adequately direct the capabilities of strong models? In a recent and somewhat surprising work, Burns et al. (2023) empirically demonstrated that when strong models (like GPT-4) are finetuned using labels generated by weak supervisors (like GPT-2), the strong models outperform their weaker counterparts---a phenomenon they term .In this work, we present a theoretical framework for understanding weak-to-strong generalization. Specifically, we show that the improvement in performance achieved by strong models over their weaker counterparts is quantified by the incurred by the strong model on labels generated by the weaker model. Our theory reveals several curious algorithmic insights. For instance, we can predict the amount by which the strong model will improve over the weak model, and also choose among different weak models to train the strong model, based on its misfit error.

Large Language Models (LLMs) are becoming a prominent generative AI tool, where the user enters a query and the LLM generates an answer. To reduce harm and misuse, efforts have been made to align these LLMs to human values using advanced training techniques such as Reinforcement Learning from Human Feedback (RLHF). However, recent studies have highlighted the vulnerability of LLMs to adversarial jailbreak attempts aiming at subverting the embedded safety guardrails.

Reinforcement Learning from Human Feedback (RLHF)has been crucial to the recent success of Large Language Models (LLMs), however it is often a complex and brittle process. In the classical RLHF framework, a reward model is first trained to represent human preferences, which is in turn used by an online reinforcement learning (RL) algorithm to optimized the LLM. A prominent issue with such methods is reward over-optimization or reward hacking, where the performance as measured by the learned proxy reward model increases, but the true model quality plateaus or even deteriorates. Direct Alignment Algorithms (DDAs), such as Direct Preference Optimization (DPO) have emerged as alternatives to the classical RLHF pipeline. However, despite not training a separate proxy reward model or using RL, they still commonly deteriorate from over-optimization. While the so-called reward hacking phenomenon is not well-defined for DAAs, we still uncover similar trends: at higher KL-budgets, DAA algorithms exhibit similar degradation patters to their classic RLHF counterparts. In particular, we find that DAA methods deteriorate not only across a wide range of KL-budgets, but also often before even a single epoch of the dataset is completed. Through extensive empirical experimentation this work formulates the reward over-optimization or hacking problem for DAAs and explores its consequences across objectives, training regimes, and model scales.

We introduce SELF-DISCOVER, a general framework for LLMs to self-discover the task-intrinsic reasoning structures to tackle complex reasoning problems that are challenging for typical prompting methods. Core to the framework is a self-discovery process where LLMs select multiple atomic reasoning modules such as critical thinking and step-by-step thinking, and compose them into an explicit reasoning structure for LLMs to follow during decoding. SELF-DISCOVER substantially improves GPT-4 and PaLM 2's performance on challenging reasoning benchmarks such as BigBench-Hard, grounded agent reasoning, and MATH, by as much as 32% compared to Chain of Thought (CoT). Furthermore, SELF-DISCOVER outperforms inference-intensive methods such as CoT-Self-Consistency by more than 20%, while requiring 10-40x fewer inference compute. Finally, we show that the self-discovered reasoning structures are universally applicable across model families: from PaLM 2-L to GPT-4, and from GPT-4 to Llama2, and share commonalities with human reasoning patterns.