arXiv subjects

Agrin Hilmkil

Publications and source records attributed to Agrin Hilmkil.

15 recordsLinked to original sources

arXiv2026

Amortized Inference of Causal Models via Conditional Fixed-Point Iterations

Structural Causal Models (SCMs) offer a principled framework to reason about interventions and support out-of-distribution generalization, which are key goals in scientific discovery. However, the task of learning SCMs from observed data poses formidable challenges, and often requires training a separate model for each dataset. In this work, we propose an amortized inference framework that trains a single model to predict the causal mechanisms of SCMs conditioned on their observational data and causal graph. We first use a transformer-based architecture for amortized learning of dataset embeddings, and then extend the Fixed-Point Approach (FiP) to infer the causal mechanisms conditionally on their dataset embeddings. As a byproduct, our method can generate observational and interventional data from novel SCMs at inference time, without updating parameters. Empirical results show that our amortized procedure performs on par with baselines trained specifically for each dataset on both in and out-of-distribution problems, and also outperforms them in scarce data regimes.

cs.LG↗

arXiv2026

Latent-Y: A Lab-Validated Autonomous Agent for De Novo Drug Design

Drug discovery relies on iterative expert workflows that are slow to parallelize and difficult to scale. Here we introduce Latent-Y, an AI agent that autonomously executes complete antibody design campaigns from text prompts, covering literature review, target analysis, epitope identification, candidate design, computational validation, and selection of lab-ready sequences. Latent-Y is integrated into the Latent Labs Platform, where it operates in the same environment as drug-discovery experts with access to bioinformatics tools, biological databases, and scientific literature. The agent can run fully autonomously end-to-end, or collaboratively, where researchers review progress, provide feedback, and direct subsequent steps. Candidate antibodies are generated using Latent-X2, our frontier generative model for drug-like antibody design. We demonstrate the agent's capability across three distinct campaign types: epitope discovery guided by therapeutic specifications, cross-species binder design, and autonomous design from a scientific publication targeting human transferrin receptor for blood-brain barrier crossing. Across nine targets, Latent-Y produced lab-confirmed nanobody binders against six, achieving a 67% target-level success rate with binding affinities reaching the single-digit nanomolar range, without human filtering or intervention. In user studies, experts working with Latent-Y completed design campaigns 56 times faster than independent expert time estimates, compressing weeks of work into hours. Because Latent-X2 is a general-purpose atomic-level model for biologics design, the same agent architecture naturally extends to macrocyclic peptide and mini-binder design campaigns, broadening autonomous discovery across therapeutic modalities. Latent-Y is available to selected partners at https://platform.latentlabs.com.

q-bio.BM↗

arXiv2025

Drug-like antibodies with low immunogenicity in human panels designed with Latent-X2

Drug discovery has long sought computational systems capable of designing drug-like molecules directly: developable and non-immunogenic from the start. Here we introduce Latent-X2, a frontier generative model that achieves this goal through zero-shot design of antibodies with strong binding affinities, drug-like properties, and, for the first time for any de novo generated antibody, confirmed low immunogenicity in human donor panels. Latent-X2 is an all-atom model conditioned on target structure, epitope specification, and optional antibody framework, jointly generating sequences and structures while modelling the bound complex. Testing only 4 to 24 designs per target in each modality, we successfully generated VHH and scFv antibodies against 9 of 18 evaluated targets, achieving a 50% target-level success rate with picomolar to nanomolar binding affinities. Designed molecules exhibit developability profiles that match or exceed those of approved antibody therapeutics, including expression yield, aggregation propensity, polyreactivity, hydrophobicity, and thermal stability, without optimization, filtering, or selection. In the first immunogenicity assessment of any AI-generated antibody, representative de novo VHH binders targeting TNFL9 exhibit both potent target engagement and low immunogenicity across T-cell proliferation and cytokine release assays. The model generalizes beyond antibodies: against K-Ras, long considered undruggable, we generated macrocyclic peptide binders competitive with trillion-scale mRNA display screens. These properties emerge directly from the model, demonstrating the therapeutic viability of zero-shot molecular design, now available without AI infrastructure or coding expertise at https://platform.latentlabs.com.

q-bio.BM↗

arXiv2025

Latent-X: An Atom-level Frontier Model for De Novo Protein Binder Design

Traditional drug discovery relies on rounds of screening millions of candidate molecules with low success rates, making drug discovery time and resource intensive. To overcome this screening bottleneck, we introduce Latent-X, an all-atom protein design model that enables a new paradigm of precision AI design. Given a target protein epitope, Latent-X jointly generates the all atom structure and sequence of the protein binder and target, directly modelling the non-covalent interactions essential for specific binding. We demonstrate its efficacy across two therapeutically relevant modalities through extensive wet lab experiments, testing as few as 30-100 designs per target. For macrocyclic peptides, Latent-X achieves experimental hit rates exceeding 90% on all evaluated benchmark targets. For mini-binders, it consistently produces potent candidates against all evaluated benchmark targets, with binding affinities reaching the low nanomolar and picomolar range - comparable to those of approved therapeutics - whilst also being highly specific in mammalian display. In direct comparisons with the state-of-the-art models AlphaProteo, RFdiffusion and RFpeptides under identical conditions demonstrates, Latent-X generates binders with higher hit rates and better binding affinities, and uniquely creates structurally diverse binders, including complex beta-sheet folds. Its end-to-end process is an order of magnitude faster than existing multi-step computational pipelines. By drastically improving the efficiency and success rate of de novo design, Latent-X represents a significant advance towards push-button biologics discovery and a valuable tool for protein engineers. Latent-X is available at https://platform.latentlabs.com, enabling users to reliably generate de novo binders without AI infrastructure or coding.

q-bio.BM↗

arXiv2024

A Fixed-Point Approach for Causal Generative Modeling

We propose a novel formalism for describing Structural Causal Models (SCMs) as fixed-point problems on causally ordered variables, eliminating the need for Directed Acyclic Graphs (DAGs), and establish the weakest known conditions for their unique recovery given the topological ordering (TO). Based on this, we design a two-stage causal generative model that first infers in a zero-shot manner a valid TO from observations, and then learns the generative SCM on the ordered variables. To infer TOs, we propose to amortize the learning of TOs on synthetically generated datasets by sequentially predicting the leaves of graphs seen during training. To learn SCMs, we design a transformer-based architecture that exploits a new attention mechanism enabling the modeling of causal structures, and show that this parameterization is consistent with our formalism. Finally, we conduct an extensive evaluation of each method individually, and show that when combined, our model outperforms various baselines on generated out-of-distribution problems. The code is available on \href{https://github.com/microsoft/causica/tree/main/research_experiments/fip}{Github}.

cs.LG↗

arXiv2024

Pyramid Vector Quantization for LLMs

Recent works on compression of large language models (LLM) using quantization considered reparameterizing the architecture such that weights are distributed on the sphere. This demonstratively improves the ability to quantize by increasing the mathematical notion of coherence, resulting in fewer weight outliers without affecting the network output. In this work, we aim to further exploit this spherical geometry of the weights when performing quantization by considering Pyramid Vector Quantization (PVQ) for large language models. Arranging points evenly on the sphere is notoriously difficult, especially in high dimensions, and in case approximate solutions exists, representing points explicitly in a codebook is typically not feasible due to its additional memory cost. Instead, PVQ uses a fixed integer lattice on the sphere by projecting points onto the 1-sphere, which allows for efficient encoding and decoding without requiring an explicit codebook in memory. To obtain a practical algorithm, we propose to combine PVQ with scale quantization for which we derive theoretically optimal quantizations, under empirically verified assumptions. Further, we extend pyramid vector quantization to use Hessian information to minimize quantization error under expected feature activations, instead of only relying on weight magnitudes. Experimentally, we achieves state-of-the-art quantization performance with pareto-optimal trade-off between performance and bits per weight and bits per activation, compared to compared methods. On weight-only, we find that we can quantize a Llama-3 70B model to 3.25 bits per weight and retain 98\% accuracy on downstream tasks.

cs.LG↗

arXiv2024

AVID: Adapting Video Diffusion Models to World Models

Large-scale generative models have achieved remarkable success in a number of domains. However, for sequential decision-making problems, such as robotics, action-labelled data is often scarce and therefore scaling-up foundation models for decision-making remains a challenge. A potential solution lies in leveraging widely-available unlabelled videos to train world models that simulate the consequences of actions. If the world model is accurate, it can be used to optimize decision-making in downstream tasks. Image-to-video diffusion models are already capable of generating highly realistic synthetic videos. However, these models are not action-conditioned, and the most powerful models are closed-source which means they cannot be finetuned. In this work, we propose to adapt pretrained video diffusion models to action-conditioned world models, without access to the parameters of the pretrained model. Our approach, AVID, trains an adapter on a small domain-specific dataset of action-labelled videos. AVID uses a learned mask to modify the intermediate outputs of the pretrained model and generate accurate action-conditioned videos. We evaluate AVID on video game and real-world robotics data, and show that it outperforms existing baselines for diffusion model adaptation.1 Our results demonstrate that if utilized correctly, pretrained video models have the potential to be powerful tools for embodied AI.

cs.CV↗

arXiv2024

Towards Causal Foundation Model: on Duality between Causal Inference and Attention

Foundation models have brought changes to the landscape of machine learning, demonstrating sparks of human-level intelligence across a diverse array of tasks. However, a gap persists in complex tasks such as causal inference, primarily due to challenges associated with intricate reasoning steps and high numerical precision requirements. In this work, we take a first step towards building causally-aware foundation models for treatment effect estimations. We propose a novel, theoretically justified method called Causal Inference with Attention (CInA), which utilizes multiple unlabeled datasets to perform self-supervised causal learning, and subsequently enables zero-shot causal inference on unseen tasks with new data. This is based on our theoretical results that demonstrate the primal-dual connection between optimal covariate balancing and self-attention, facilitating zero-shot causal inference through the final layer of a trained transformer-type architecture. We demonstrate empirically that CInA effectively generalizes to out-of-distribution datasets and various real-world datasets, matching or even surpassing traditional per-dataset methodologies. These results provide compelling evidence that our method has the potential to serve as a stepping stone for the development of causal foundation models.

cs.LG↗

arXiv2024

The Essential Role of Causality in Foundation World Models for Embodied AI

Recent advances in foundation models, especially in large multi-modal models and conversational agents, have ignited interest in the potential of generally capable embodied agents. Such agents will require the ability to perform new tasks in many different real-world environments. However, current foundation models fail to accurately model physical interactions and are therefore insufficient for Embodied AI. The study of causality lends itself to the construction of veridical world models, which are crucial for accurately predicting the outcomes of possible interactions. This paper focuses on the prospects of building foundation world models for the upcoming generation of embodied agents and presents a novel viewpoint on the significance of causality within these. We posit that integrating causal considerations is vital to facilitating meaningful physical interactions with the world. Finally, we demystify misconceptions about causality in this context and present our outlook for future research.

cs.AI↗

arXiv2023

Learned Causal Method Prediction

For a given causal question, it is important to efficiently decide which causal inference method to use for a given dataset. This is challenging because causal methods typically rely on complex and difficult-to-verify assumptions, and cross-validation is not applicable since ground truth causal quantities are unobserved. In this work, we propose CAusal Method Predictor (CAMP), a framework for predicting the best method for a given dataset. To this end, we generate datasets from a diverse set of synthetic causal models, score the candidate methods, and train a model to directly predict the highest-scoring method for that dataset. Next, by formulating a self-supervised pre-training objective centered on dataset assumptions relevant for causal inference, we significantly reduce the need for costly labeled data and enhance training efficiency. Our strategy learns to map implicit dataset properties to the best method in a data-driven manner. In our experiments, we focus on method prediction for causal discovery. CAMP outperforms selecting any individual candidate method and demonstrates promising generalization to unseen semi-synthetic and real-world benchmarks.

cs.LG↗

arXiv2023

Understanding Causality with Large Language Models: Feasibility and Opportunities

We assess the ability of large language models (LLMs) to answer causal questions by analyzing their strengths and weaknesses against three types of causal question. We believe that current LLMs can answer causal questions with existing causal knowledge as combined domain experts. However, they are not yet able to provide satisfactory answers for discovering new knowledge or for high-stakes decision-making tasks with high precision. We discuss possible future directions and opportunities, such as enabling explicit and implicit causal modules as well as deep causal-aware LLMs. These will not only enable LLMs to answer many different types of causal questions for greater impact but also enable LLMs to be more trustworthy and efficient in general.

cs.LG↗

arXiv2023

Causal Reasoning in the Presence of Latent Confounders via Neural ADMG Learning

Latent confounding has been a long-standing obstacle for causal reasoning from observational data. One popular approach is to model the data using acyclic directed mixed graphs (ADMGs), which describe ancestral relations between variables using directed and bidirected edges. However, existing methods using ADMGs are based on either linear functional assumptions or a discrete search that is complicated to use and lacks computational tractability for large datasets. In this work, we further extend the existing body of work and develop a novel gradient-based approach to learning an ADMG with non-linear functional relations from observational data. We first show that the presence of latent confounding is identifiable under the assumptions of bow-free ADMGs with non-linear additive noise models. With this insight, we propose a novel neural causal model based on autoregressive flows for ADMG learning. This not only enables us to determine complex causal structural relationships behind the data in the presence of latent confounding, but also estimate their functional relationships (hence treatment effects) simultaneously. We further validate our approach via experiments on both synthetic and real-world datasets, and demonstrate the competitive performance against relevant baselines.

cs.LG↗

arXiv2021

Perceiving Music Quality with GANs

Several methods have been developed to assess the perceptual quality of audio under transforms like lossy compression. However, they require paired reference signals of the unaltered content, limiting their use in applications where references are unavailable. This has hindered progress in audio generation and style transfer, where a no-reference quality assessment method would allow more reproducible comparisons across methods. We propose training a GAN on a large music library, and using its discriminator as a no-reference quality assessment measure of the perceived quality of music. This method is unsupervised, needs no access to degraded material and can be tuned for various domains of music. In a listening test with 448 human subjects, where participants rated professionally produced music tracks degraded with different levels and types of signal degradations such as waveshaping distortion and low-pass filtering, we establish a dataset of human rated material. By using the human rated dataset we show that the discriminator score correlates significantly with the subjective ratings, suggesting that the proposed method can be used to create a no-reference musical audio quality assessment measure.

cs.SD↗

arXiv2021

Scaling Federated Learning for Fine-tuning of Large Language Models

Federated learning (FL) is a promising approach to distributed compute, as well as distributed data, and provides a level of privacy and compliance to legal frameworks. This makes FL attractive for both consumer and healthcare applications. While the area is actively being explored, few studies have examined FL in the context of larger language models and there is a lack of comprehensive reviews of robustness across tasks, architectures, numbers of clients, and other relevant factors. In this paper, we explore the fine-tuning of Transformer-based language models in a federated learning setting. We evaluate three popular BERT-variants of different sizes (BERT, ALBERT, and DistilBERT) on a number of text classification tasks such as sentiment analysis and author identification. We perform an extensive sweep over the number of clients, ranging up to 32, to evaluate the impact of distributed compute on task performance in the federated averaging setting. While our findings suggest that the large sizes of the evaluated models are not generally prohibitive to federated training, we found that the different models handle federated averaging to a varying degree. Most notably, DistilBERT converges significantly slower with larger numbers of clients, and under some circumstances, even collapses to chance level performance. Investigating this issue presents an interesting perspective for future research.

cs.LG↗

arXiv2018

Towards Machine Learning on data from Professional Cyclists

Professional sports are developing towards increasingly scientific training methods with increasing amounts of data being collected from laboratory tests, training sessions and competitions. In cycling, it is standard to equip bicycles with small computers recording data from sensors such as power-meters, in addition to heart-rate, speed, altitude etc. Recently, machine learning techniques have provided huge success in a wide variety of areas where large amounts of data (big data) is available. In this paper, we perform a pilot experiment on machine learning to model physical response in elite cyclists. As a first experiment, we show that it is possible to train a LSTM machine learning algorithm to predict the heart-rate response of a cyclist during a training session. This work is a promising first step towards developing more elaborate models based on big data and machine learning to capture performance aspects of athletes.

cs.LG↗