Universität Bonn

Hertz-Chair for Artificial Intelligence and Neurosciene

Seminar "Biological Intelligence " (MA-INF 4329)

Humans and other animals outperform artificial agents in various tasks and domains.

This includes but is not limited to: learning and planning in unstructured domains; learning from sparse data, observation, and play; generalisation and transfer; causal reasoning; intuitive physics and psychology; language use; any time planning; continuous planning; spatial navigation; dynamic memory and active forgetting.

This seminar provides background on some of the underlying biological skills and computational theories that seek to explain them. We will discuss implications for designing and/or constraining artificial agents.

Meeting Times and Location: 

Mondays, 16:00 - 17.30 (s.t.)

Seminarraum 1.047

Institute for Computer Science (Informatik III)

Friedrich-Hirzebruch-Allee 8

53117 Bonn

Registration 

If you are interested in participating in the seminar, please send an e-mail to Anja Menke (caian.office@uni-bonn.de) and come to the first meeting.

If you would like to participate in the seminar but cannot attend the first meeting, please send us an email before the first meeting.

Schedule

Background presentation and assignment of topics.

This session will take place online due to the scheduled bomb disposal on campus.

Using cognitive psychology to understand GPT-3.

Reading:

Binz & Schulz [Link]

AGENT: A Benchmark for Core Psychological ReasoningAGENT: A Benchmark for Core Psychological Reasoning

Reading:

Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models.

Reading:

Srivastava et al. [Link]

Navigation Turing Test (NTT): Learning to Evaluate Human-Like Navigation.

Reading:

Devlin et al. [Link]

No class.

A Benchmark for Core Psychological Reasoning.

Theories of Error Back-Propagation in the Brain.

Reading: 

Whittington & Bogasz [Link]

Neural population geometry and symmetries.

Reading:

Chung & Abbot, Higgins et al. [Link1] [Link2]

Generalisation and bias in image recognition.

Reading: 

Gerhos et al. Link 1 Link 2

A developmental approach to machine-learning.

Reading:

Smith & Slone, Lake & Brendon [Link 1] [Link 2]

Human-Level Reinforcement Learning through Theory-Based Modeling, Exploration, and Planning.

Reading:

Tsividis et al. [Link]

Human-like systematic generalization through a meta-learning neural network.

Reading:

Lake & Baroni [Link]

Theories of Error Back-Propagation in the Brain.

Reading:

Wichmann & Geirhos [Link]


Contact

Avatar Menke

Anja Menke

Administrative Assistant (Geschäftszimmer)

Am Propsthof 49

53121 Bonn

Wird geladen