Machine Intelligence Lab

Research lab @UofT Reverse-engineering intelligence. Building intelligent machines. Director @DrLaschowski

[Machine Intelligence Lab @LaschowskiLab]

Meet Brandon Wong, an MSc student @UofTCompSci studying representation learning for neural decoding.

His current research explores how diffusion latents affect downstream learning.

[Brokoslaw Laschowski @DrLaschowski]

Invited talk at @MSFTResearch on building computational models of intelligence. Thanks for having me.

@UofT @LaschowskiLab

[Brokoslaw Laschowski @DrLaschowski]

Proud to lead the @UofT summer research program in computer science for students from Ukraine.

@UofTCompSci @VectorInst

[Brokoslaw Laschowski @DrLaschowski]

Our mission: reverse-engineer intelligence. Follow Machine Intelligence Lab for updates.

@UofT @LaschowskiLab

[Dmytro Kuzmenko @kuzmenko_dmytro]

so it begins @eccvconf

[Brokoslaw Laschowski @DrLaschowski]

@UofTEngineering @ChemEng_UofT @UofT @uoftmie Come learn about our research on reverse-engineering intelligence and the physics of learning and neural computation.

@LaschowskiLab

[Sofiya Zbaranska @neur1s]

I heard you need help with some bugs

[Brokoslaw Laschowski @DrLaschowski]

Our new study on bidirectional representational alignment between biological and artificial neural networks.

@UofT @LaschowskiLab https://t.co/DXzxWCyZjl

[Brokoslaw Laschowski @DrLaschowski]

Are you a @UofT student in computer science, math, physics, or engineering? Want research experience?

We’re recruiting students to explore theory and algorithms for inverse reinforcement learning. Email me your resume and transcript.

[Brokoslaw Laschowski @DrLaschowski]

I started in neuroscience. Then spent several years building intelligent machines. Now I study general principles of learning and intelligence.

@UofT @LaschowskiLab

[Brokoslaw Laschowski @DrLaschowski]

Proud professor. The next-generation of machine learning researchers from Ukraine.

@UofT @LaschowskiLab

[Brokoslaw Laschowski @DrLaschowski]

@kuzmenko_dmytro Full-circle. When your students go on to supervise their own students. Congrats Dima on the @eccvconf papers. Proud of you.

[Machine Intelligence Lab @LaschowskiLab]

https://t.co/apk2pLhe31

[Machine Intelligence Lab @LaschowskiLab]

Valence-driven memory prioritization in machines by Aditya Rajeev, Sofiya Zbaranska, Sheena Josselyn, and Brokoslaw Laschowski.

@UofT

[Machine Intelligence Lab @LaschowskiLab]

https://t.co/AIsfiTa7i1

[Machine Intelligence Lab @LaschowskiLab]

Meet Brandon Wong, an MSc student @UofTCompSci studying representation learning for neural decoding.

His current research explores how diffusion latents affect downstream learning.

[Machine Intelligence Lab @LaschowskiLab]

https://t.co/WXUOcZv7Ac

[Machine Intelligence Lab @LaschowskiLab]

Bidirectional representational alignment between brains and machines by Samuel Kostousov, Abhin Kaushik, and Brokoslaw Laschowski.

@UofT

[Brokoslaw Laschowski @DrLaschowski]

Introducing our framework for interpreting reward functions recovered by inverse reinforcement learning.

https://t.co/ThJOmWCcqW @UofT @LaschowskiLab

[Brokoslaw Laschowski @DrLaschowski]

Applications are now open for the @VectorInst Distinguished Postdoctoral Fellowships. Come join our world-class machine learning research community.

Apply by August 31: https://t.co/rOUUUNgqm7

[Brokoslaw Laschowski @DrLaschowski]

@SergeyStavisky Thanks for your feedback. We use "domain" in the machine learning sense, where different users, tasks, sessions, and devices correspond to different data distributions. In that sense, we view cross-user transfer as a special case of cross-domain adaptation.

[Brokoslaw Laschowski @DrLaschowski]

@SergeyStavisky We see these as complementary. Scaling emphasizes data quantity, whereas source-domain adaptation emphasizes data selection. Both are important for transfer performance and data efficiency in neural decoding.

[Brokoslaw Laschowski @DrLaschowski]

@SergeyStavisky Interesting. Is scaling alone sufficient for cross-domain decoding? Not all source domains are equally informative, and indiscriminate scaling can induce negative transfer. Selective source-domain adaptation can improve both transfer and data efficiency. https://t.co/l9P7tOd7es