Speakers
Programmed at ML, in alphabetical order.
Agenda
Selected from 20 sessions.
- Thursday / 8 October 08:15 - 08:30Student Research Workshop
Interpretability-Guided Soft Pruning of Attention Heads in Vision Transformers
Vision foundation models, such as DINOv2, learn highly expressive representations but rely on massive, opaque architectures that demand substantial computational power and memory. …
- Thursday / 8 October 08:35 - 08:50Student Research Workshop
Diverse Subset Selection at Scale with Maximum Diversity Picking
Data scarcity is a common problem in machine learning. However, researchers often face the opposite situation, where processing the entire dataset is too expensive or even infeasib…
- Thursday / 8 October 08:55 - 09:10Student Research Workshop
Goal Conditioned RL to Abstract Solved States via Relative Contrastive Reinforcement Learning
Goal-Conditioned Reinforcement Learning (GCRL) has been proposed as an effective way to solve tasks with sparse rewards, offering an alternative to tedious reward engineering. Intu…
- Thursday / 8 October 09:15 - 09:30Student Research Workshop
Activity is in the Activations: Inference-Time Steering of Peptide Models for Antimicrobial Design
While recent generative diffusion models offer intuitive control over complex outputs, transferring this steerability to targeted scientific discovery remains a significant challen…
- Thursday / 8 October 10:00 - 10:15Student Research Workshop
Conditioned Activation Transport for T2I Safety Steering
Despite their impressive capabilities, current Text-to-Image (T2I) models remain prone to generating unsafe and toxic content. While activation steering offers a promising inferenc…
- Thursday 8 October, 10:00 - 10:15Student Research Workshop
Conditioned Activation Transport for T2I Safety Steering
Despite their impressive capabilities, current Text-to-Image (T2I) models remain prone to generating unsafe and toxic content. While activation steering offers a promising inferenc…
- Thursday / 8 October 10:20 - 10:35Student Research Workshop
Drobage: Closing the Perception–Grasp Loop for Manipulation and Navigation in Unstructured Outdoor Environments
Outdoor cleaning robots exist, and nearly all of them are blind: sweepers and suction units displace whatever they pass over, with no representation of an individual object and no …
- Thursday 8 October, 10:20 - 10:35Student Research Workshop
Drobage: Closing the Perception–Grasp Loop for Manipulation and Navigation in Unstructured Outdoor Environments
Outdoor cleaning robots exist, and nearly all of them are blind: sweepers and suction units displace whatever they pass over, with no representation of an individual object and no …
- Thursday / 8 October 10:40 - 10:55Student Research Workshop
GS-JEPA: A Self-Supervised Approach to Learning Gaussian Splat Representations Using Joint Embedding Predictive Architecture
Current self-supervised learning methods for Gaussian splat object representations primarily rely on reconstruction-based pretraining, where the model is optimized to reconstruct i…
An audience of 8 companies, parsed from the program.
Eight companies in attendance.
The numbers below are derived from the speakers, sponsors, and exhibitors on this page, cross-referenced into one ledger. They are the only thing here that 10times.com cannot tell you.