We use cookies to analyze the browsing and usage of our website and to personalize your experience. You can disable these technologies at any time, but this may limit certain functionalities of the site. Read our Privacy Policy for more information.
Setting cookies
You can enable and disable the types of cookies you wish to accept. However certain choices you make could affect the services offered on our sites (e.g. suggestions, personalised ads, etc.).
Essential cookies
These cookies are necessary for the operation of the site and cannot be deactivated. (Still active)
Analytics cookies
Do you accept the use of cookies to measure the audience of our sites?
Multimedia Player
Do you accept the use of cookies to display and allow you to watch the video content hosted by our partners (YouTube, etc.)?
Publications
SafeArena: Evaluating the Safety of Autonomous Web Agents
LLM-based agents are becoming increasingly proficient at solving web-based tasks. With this capability comes a greater risk of misuse for ma… (see more)licious purposes, such as posting misinformation in an online forum or selling illicit substances on a website. To evaluate these risks, we propose SafeArena, the first benchmark to focus on the deliberate misuse of web agents. SafeArena comprises 250 safe and 250 harmful tasks across four websites. We classify the harmful tasks into five harm categories -- misinformation, illegal activity, harassment, cybercrime, and social bias, designed to assess realistic misuses of web agents. We evaluate leading LLM-based web agents, including GPT-4o, Claude-3.5 Sonnet, Qwen-2-VL 72B, and Llama-3.2 90B, on our benchmark. To systematically assess their susceptibility to harmful tasks, we introduce the Agent Risk Assessment framework that categorizes agent behavior across four risk levels. We find agents are surprisingly compliant with malicious requests, with GPT-4o and Qwen-2 completing 34.7% and 27.3% of harmful requests, respectively. Our findings highlight the urgent need for safety alignment procedures for web agents. Our benchmark is available here: https://safearena.github.io
Diffusion Probabilistic Models (DPMs) are powerful generative models that have achieved unparalleled success in a number of generative tasks… (see more). In this work, we aim to build inductive biases into the training and sampling of diffusion models to better accommodate the target distribution of the data to model. For topologically structured data, we devise a frequency-based noising operator to purposefully manipulate, and set, these inductive biases. We first show that appropriate manipulations of the noising forward process can lead DPMs to focus on particular aspects of the distribution to learn. We show that different datasets necessitate different inductive biases, and that appropriate frequency-based noise control induces increased generative performance compared to standard diffusion. Finally, we demonstrate the possibility of ignoring information at particular frequencies while learning. We show this in an image corruption and recovery task, where we train a DPM to recover the original target distribution after severe noise corruption.
This paper presents a practical application of Relative Trajectory Balance (RTB), a recently introduced off-policy reinforcement learning (R… (see more)L) objective that can asymptotically solve Bayesian inverse problems optimally. We extend the original work by using RTB to train conditional diffusion model posteriors from pretrained unconditional priors for challenging linear and non-linear inverse problems in vision, and science. We use the objective alongside techniques such as off-policy backtracking exploration to improve training. Importantly, our results show that existing training-free diffusion posterior methods struggle to perform effective posterior inference in latent space due to inherent biases.
Personalized machine learning models have gained significant importance in various domains, including healthcare. However, designing efficie… (see more)nt personalized models remains a challenge. Traditional approaches often involve training multiple sub-models for different population sub-groups, which can be costly and does not always guarantee improved performance across all sub-groups. This paper presents a novel approach to improving model performance at the sub-group level by leveraging bias and training a joint model. Our method involves a two-step process: first, we train a model to predict group attributes, and then we use this model to learn data-dependent biases to modulate a second model for diagnosis prediction. Our results demonstrate that this joint architecture achieves consistent performance gains across all sub-groups in the Heart dataset. Furthermore, in the mortality dataset, it improves performance in two of the four sub-groups. A comparison of our method with the traditional decoupled personalization method demonstrated a greater performance gain in the sub-groups with less harm. This approach offers a more effective and scalable solution for personalization of models, which could have positive impact in healthcare and other areas that require predictive models which take sub-group information into account.
Clinical prediction models are often personalized to target heterogeneous sub-groups by using demographic attributes such as race and gender… (see more) to train the model. Traditional personalization approaches involve using demographic attributes in input features or training multiple sub-models for different population subgroups (decoupling model). However, these methods often harm the performance at the
subgroup level compared to non-personalized models. This paper presents a novel personalization method to improve model performance at the sub-group level. Our method involves a two-step process: first, we train a model to predict group attributes, and then we use this model to learn data-dependent biases to modulate a second model for diagnosis prediction. Our results demonstrate that this joint architecture achieves consistent performance gains across all sub-groups in the Heart dataset. Furthermore, in the mortality dataset, it improves performance in two of the four sub-groups. A comparison of our method with the traditional decoupled personalization method demonstrated a greater performance gain in the sub-groups with less harm. This approach offers a more effective and scalable solution for personalized models, which could have a positive impact in healthcare and other areas that require predictive models that take sub-group information into
account.
Proteins perform diverse biological functions, governed by the intricate relationship between their sequence and three-dimensional structure… (see more). While protein language models (PLMs) have demonstrated remarkable success in functional annotation and structure prediction, their potential for sequence-structure co-design remains underexplored. This limitation arises from pre-training objectives that favor masked token prediction over generative modeling. In this work, we systematically explore sampling strategies to enhance the generative capabilities of PLMs for co-design. Notably, we introduce a ranked iterative decoding with re-masking scheme, enabling PLMs to generate sequences and structures more effectively. Benchmarking ESM3 across multiple scales, we demonstrate that using PLMs effectively at sampling time for co-design tasks can outperform specialized architectures that lack comparable scaling properties. Our work advances the field of computational protein design by equipping PLMs with robust generative capabilities tailored to sequence-structure interdependence.
Proteins perform diverse biological functions, governed by the intricate relationship between their sequence and three-dimensional structure… (see more). While protein language models (PLMs) have demonstrated remarkable success in functional annotation and structure prediction, their potential for sequence-structure co-design remains underexplored. This limitation arises from pre-training objectives that favor masked token prediction over generative modeling. In this work, we systematically explore sampling strategies to enhance the generative capabilities of PLMs for co-design. Notably, we introduce a ranked iterative decoding with re-masking scheme, enabling PLMs to generate sequences and structures more effectively. Benchmarking ESM3 across multiple scales, we demonstrate that using PLMs effectively at sampling time for co-design tasks can outperform specialized architectures that lack comparable scaling properties. Our work advances the field of computational protein design by equipping PLMs with robust generative capabilities tailored to sequence-structure interdependence.
Federated learning is a machine learning paradigm where multiple clients collaboratively train a global model by exchanging their locally tr… (see more)ained model weights instead of raw data. In the standard setting, every client trains the local model for the same number of epochs.
We introduce ALT (Adaptive Local Training), a simple yet effective feedback mechanism that can be exploited at the client side to limit unnecessary and degrading computations. ALT dynamically adjusts the number of training epochs for each client based on the similarity between their local representations and the global one, ensuring that well-aligned clients can train longer without experiencing client drift. We evaluated ALT on federated partitions of the CIFAR-10 and Tiny-ImageNet datasets, demonstrating its effectiveness in improving model convergence and stability.
Federated Learning is a machine learning paradigm where multiple clients collaboratively train a global model by exchanging their locally tr… (see more)ained model weights instead of raw data.
In the standard setting, every client trains the local model for the same number of epochs.
We introduce ALT (Adaptive Local Training), a simple yet effective feedback mechanism that could be introduced at the client side to limit unnecessary and degrading computations.
ALT dynamically adjusts the number of training epochs for each client based on the similarity between their local representations and the global one, ensuring that well-aligned clients can train longer without experiencing client drift. We evaluated ALT on federated partitions of the CIFAR-10 and TinyImageNet datasets, demonstrating its effectiveness in improving model convergence and stability.
Aligning visual features with language embeddings is a key challenge in vision-language models (VLMs). The performance of such models hinges… (see more) on having a good connector that maps visual features generated by a vision encoder to a shared embedding space with the LLM while preserving semantic similarity. Existing connectors, such as multilayer perceptrons (MLPs), often produce out-of-distribution or noisy inputs, leading to misalignment between the modalities. In this work, we propose a novel vision-text alignment method, AlignVLM, that maps visual features to a weighted average of LLM text embeddings. Our approach leverages the linguistic priors encoded by the LLM to ensure that visual features are mapped to regions of the space that the LLM can effectively interpret. AlignVLM is particularly effective for document understanding tasks, where scanned document images must be accurately mapped to their textual content. Our extensive experiments show that AlignVLM achieves state-of-the-art performance compared to prior alignment methods. We provide further analysis demonstrating improved vision-text feature alignment and robustness to noise.
