The upcoming meeting, taking place on November 10 at Mila, will explore how we can collectively develop, govern, and deploy high-performing, reliable, and secure agentic systems by connecting academic researchers, industry experts, and practitioners.
Join us on the Venture Scientist Bootcamp, a full time, 4-month incubator at Mila, built specifically for deep tech founders with elite STEM backgrounds. You have until November 21, 2026 to apply!
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
Imagining Goals for Policy Generalization with Geometric Horizon Models
Training from offline data has allowed for substantial progress in domains such as robotics, leading to general-purpose policies that can be… (see more) easily applied zero-shot or efficiently finetuned for downstream tasks. However, training policies on offline data can lead to poor generalization, both due to the choice of modeling objective and from learning from a static dataset. In this work, we focus on the challenging task of zero-shot goal generalization, where a policy is evaluated on unseen tasks that require reusing its existing knowledge (compositional generalization). An avenue for improving a policy's generalization is through generating new experience through world models; however, such generation has proven difficult for longer horizons. Thus, to alleviate this issue, we propose TD-Aug, sampling from a geometric horizon model, which allows for directly imagining novel outcomes that can be achieved through composing existing knowledge. We demonstrate that training on these future outcomes as goals for goal-conditioned BC and offline RL policies improves generalization in stitching-based OGBench tasks.
Many Blind and Low-Vision (BLV) people rely on guide dogs for moment-to-moment navigation, such as staying on path and avoiding obstacles an… (see more)d pedestrians. However, guide dogs are expensive to acquire and maintain (approximately \
AInimation: Animating from Prompt to AI-Generated Responses
Jiaqi Wu
Damien Masson
We explore the use of animated transitions between a prompt and an AI-generated response. After reviewing 800 examples of prompts and respon… (see more)ses, we devise a taxonomy of animated transitions for multimodal text- and image-generative models. The proposed animations include translating and morphing elements of the prompt to their final location in the response; highlighting modifications such as fixed typos; overlaying structural requirements to verify them; and displaying how a model understands references. A study shows that adding animated transitions helps users review the response: participants performed 43% better at locating elements in the response; 153% better at identifying changes; and 20% better at verifying the prompt was correctly interpreted. Our work applies to all software that integrates AI and shows that well-crafted, slower animations are preferable to instant AI responses.
Few-shot Test-Time Domain Adaptation (FSTT-DA) seeks to adapt models to novel domains using only a handful of unlabeled target samples. This… (see more) setting is more realistic than typical domain adaptation setups, which assume access to target data during source training. However, prior FSTT-DA approaches fail to effectively leverage source domain-specific knowledge, relying on shallow batch normalization updates, prompt-based methods that treat the model as a black box, or ensembling strategies that do not capture cross-domain relationships. To address these limitations, we introduce a new FSTT-DA framework that integrates LoRA fine-tuning with model merging. In our approach, separate LoRA modules are fine-tuned on CLIP's vision encoder for each source domain. Since LoRA modifies only a small fraction of the model's parameters, it retains the base model's generalized knowledge while internally learning domain-specific features. To adapt the learned knowledge to a specific target domain, we propose a hypernetwork trained via meta-learning that generates per-column merging factors to combine LoRA modules. Given a small batch of target images, the hypernetwork produces merging weights that fuse source LoRA modules into a single adapted representation. Our results demonstrate state-of-the-art performance across various domain adaptation datasets. Our code is publicly available at https://github.com/nahbois4321/DA-MergeLoRA.
Financial question answering over U.S. Securities and Exchange Commission (SEC) filings requires retrieving and synthesizing heterogeneous e… (see more)vidence dispersed across long, standardized, and highly redundant disclosures. Existing retrieval-augmented and multi-agent systems typically derive retrieval queries directly from the user's question and rank candidates by semantic similarity. Together, these choices create prior-corpus misalignment: a mismatch between model priors and the target filings' structure, terminology, and evidence standards. As a result, query generation misses corpus-specific evidence, while semantic reranking favors topically similar but evidentially invalid false-positive chunks. We propose FinSAgent, an evidence-grounded multi-agent framework that reframes SEC filing QA as corpus-aligned retrieval planning and corrects both ends with a single principle: inject corpus-side conditioning wherever model priors would otherwise dominate. FinSAgent combines (1) role-specialized agents anchored to the mandated 10-K item structure, (2) database-aware query decomposition that conditions each agent's sub-queries on a lightweight, summary-level view of the local corpus, and (3) multi-path retrieval with a learned feature-gated reranker that separates evidential validity from semantic similarity. Across five offline financial QA benchmarks, FinSAgent improves retrieval coverage and answer correctness over strong single-agent and multi-agent baselines; in a three-arm randomized online experiment with 1,000 anonymous user ratings, it also receives higher scores than baselines.
Physical AI agents acting in the real world fail differently from language models: a misjudgment of trajectory, force, or contact can have i… (see more)mmediate and potentially irreversible consequences. As video world models are increasingly deployed as the perception and dynamics layer of such agents, understanding what physical structure they actually represent internally be- comes a precondition for trustworthy deployment. Yet today these models are largely studied as black boxes, evaluated only through behavioral benchmarks. We argue that this should change: world models can be opened up, and the latent variables they encode can be measured, interpreted, and eventually controlled. We take a first step in this direction across two state-of-the-art video encoders (V-JEPA 2 and VideoMAE-v2), using layerwise probing, subspace geometry, patch-level decoding, and targeted attention ablations. We identify a sharp intermediate-depth transition we call the Physics Emergence Zone, at which physical variables become accessible. Decomposing motion into explicit variables, we find that scalar quantities such as speed and acceleration are available from early layers onwards, whereas motion direction becomes accessible only at the Physics Emergence Zone, encoded as a high-dimensional circular population code that requires coordinated multi-feature intervention to steer. These findings argue against compact, reusable latent physics state and in favor of distributed, task-specific representations supported by a shared local-attention circuit. We discuss implications for safe deploy- ment of video world models in embodied physical AI, including why intermediate-layer features, multi-feature monitors, and the local-attention circuit at the Physics Emergence Zone are natural targets for runtime verification.
2026-07-19
SPAI @ International Joint Conference on Artificial Intelligence - European Conference on Artificial Intelligence (published)
Not all training samples contribute equally to large language model fine-tuning. Selecting informative training samples can reduce the compu… (see more)tational cost while preserving downstream performance. Many existing data selection methods rely on indirect heuristics, such as data quality, diversity or reasoning trace length. However, the effectiveness of these fixed criteria is task-dependent and difficult to generalize across diverse downstream tasks. Perplexity-based data selection provides a simple and model-aware solution to estimate the sample difficulty, but existing approaches typically score the entire training sequence and ignore the difference in learning objectives of language modeling and reasoning tasks. In this paper, we propose PPL-Factory, a simple and interpretable data selection framework that combines task-aware perplexity-based scores and data budget-aware selection criteria. Experiments on GSM8K demonstrate that PPL-Factory outperforms other state-of-the-art data selection methods using only
Knowledge graph question answering (KGQA) requires navigating from topic entities to an answer several relations away. Recent methods prompt… (see more) a frontier LLM to explore the graph through a retrieval tool, but their reliance on frontier-scale inference makes them costly to deploy. We present Search-on-Graph-R1 (\sogrone{}), which internalizes this navigation into a compact 8B model through supervised fine-tuning (SFT) followed by reinforcement learning (RL). Our central idea is to scaffold a frontier teacher with each question's gold SPARQL query, so the teacher traverses a known answer-bearing path with a live \texttt{Search} tool rather than having to discover the path itself. Since every call executes against a live Freebase server, the resulting trajectories are grounded in the knowledge graph by construction. On WebQSP, CWQ, and GrailQA, \sogrone{} at 8B surpasses every frozen frontier-LLM system in our comparison and posts the strongest results on CWQ of any system we compare against. It does so using no auxiliary module at inference and no LLM judge during training. Isolating each training stage shows that SFT and RL contribute complementary gains, our approach transfers across model families, and RL learns to reach answers in fewer \texttt{Search} calls than its SFT initialization.
The stability of grid-forming (GFM) inverter-based resources (IBRs) is typically analyzed under the assumption that they are supplied by bat… (see more)teries or ideal DC sources, without accounting for variations in the DC link voltage. This paper contributes to small-signal stability analysis of photovoltaic (PV) parks operating in a two-stage GFM configuration: reserve grid-forming mode (R-GFM), which provides power reserves, and limited grid-forming mode (L-GFM), which delivers the maximum available power. A transition strategy between these modes is adopted and validated. Numerical simulations are conducted in EMTP® on a PV park connected to a weak grid under both standard and low irradiance conditions, comparing the two-stage GFM operation to the standard grid-following (GFL) control. Results show that the PV park preserves system stability in GFM modes, whereas instability occurs under GFL operation.
2026-07-18
IEEE Power & Energy Society General Meeting (published)
While cages provide a powerful reduction model to simulate elastic deformations, interactive real-time cutting of cages has been a longstand… (see more)ing challenge. Traditional solvers require costly volumetric re-meshing and global matrix updates after each modification. For a mesh-free harmonic weight computation using walk on spheres (WoS), we make a key observation: only a small fraction of walks (5-10%) is invalidated by the local cut in most practical cases. We propose a novel reuse pipeline to efficiently update only weights subject to a topological modification. Our approach leverages walk on spheres as an output sensitive and mesh-free Monte Carlo method for solving Laplacian equations, to enable localized updates of harmonic weights. By identifying and reusing unaffected walks, we minimize our re-walking updates. Our method achieves speedups of 10 × -40 × while being statistically identical to a complete WoS weight recomputation. This enables a real time simulation of interactive cutting on complex 2D and 3D cages, and is easily generalizable to other local cage editing operations.
2026-07-18
Conference on Computer Graphics and Interactive Techniques (published)
Modern power systems are increasingly dominated by inverter-based resources (IBRs), which rely on fast control systems for its power-electri… (see more)c converters. These IBRs are vulnerable to oscillating interactions with other grid components. The stability of such power systems can be studied through impedance-based stability assessment. Different impedance modelling domains offer varying trade-offs in terms of accuracy, ease of implementation, and interpretability. Each provides distinct insight into the dynamic behavior and control interactions of the system. This paper presents a concise comparative overview of impedance model types, illustrating their practical implications through representative case studies. All impedance models are measured from detailed electromagnetic transient models.
2026-07-18
IEEE Power & Energy Society General Meeting (published)