Alert button
Picture for Alexander Koller

Alexander Koller

Alert button

Simple and effective data augmentation for compositional generalization

Jan 18, 2024
Yuekun Yao, Alexander Koller

Viaarxiv icon

AutoPlanBench: : Automatically generating benchmarks for LLM planners from PDDL

Nov 16, 2023
Katharina Stein, Alexander Koller

Viaarxiv icon

Predicting generalization performance with correctness discriminators

Nov 15, 2023
Yuekun Yao, Alexander Koller

Viaarxiv icon

ADaPT: As-Needed Decomposition and Planning with Language Models

Nov 08, 2023
Archiki Prasad, Alexander Koller, Mareike Hartmann, Peter Clark, Ashish Sabharwal, Mohit Bansal, Tushar Khot

Viaarxiv icon

SLOG: A Structural Generalization Benchmark for Semantic Parsing

Oct 23, 2023
Bingzhi Li, Lucia Donatelli, Alexander Koller, Tal Linzen, Yuekun Yao, Najoung Kim

Viaarxiv icon

Closing the Curious Case of Neural Text Degeneration

Oct 02, 2023
Matthew Finlayson, John Hewitt, Alexander Koller, Swabha Swayamdipta, Ashish Sabharwal

Figure 1 for Closing the Curious Case of Neural Text Degeneration
Figure 2 for Closing the Curious Case of Neural Text Degeneration
Figure 3 for Closing the Curious Case of Neural Text Degeneration
Figure 4 for Closing the Curious Case of Neural Text Degeneration
Viaarxiv icon

Injecting a Structural Inductive Bias into a Seq2Seq Model by Simulation

Oct 01, 2023
Matthias Lindemann, Alexander Koller, Ivan Titov

Figure 1 for Injecting a Structural Inductive Bias into a Seq2Seq Model by Simulation
Figure 2 for Injecting a Structural Inductive Bias into a Seq2Seq Model by Simulation
Figure 3 for Injecting a Structural Inductive Bias into a Seq2Seq Model by Simulation
Figure 4 for Injecting a Structural Inductive Bias into a Seq2Seq Model by Simulation
Viaarxiv icon

Compositional Generalization without Trees using Multiset Tagging and Latent Permutations

May 26, 2023
Matthias Lindemann, Alexander Koller, Ivan Titov

Figure 1 for Compositional Generalization without Trees using Multiset Tagging and Latent Permutations
Figure 2 for Compositional Generalization without Trees using Multiset Tagging and Latent Permutations
Figure 3 for Compositional Generalization without Trees using Multiset Tagging and Latent Permutations
Figure 4 for Compositional Generalization without Trees using Multiset Tagging and Latent Permutations
Viaarxiv icon

What's the Meaning of Superhuman Performance in Today's NLU?

May 15, 2023
Simone Tedeschi, Johan Bos, Thierry Declerck, Jan Hajic, Daniel Hershcovich, Eduard H. Hovy, Alexander Koller, Simon Krek, Steven Schockaert, Rico Sennrich, Ekaterina Shutova, Roberto Navigli

Figure 1 for What's the Meaning of Superhuman Performance in Today's NLU?
Figure 2 for What's the Meaning of Superhuman Performance in Today's NLU?
Figure 3 for What's the Meaning of Superhuman Performance in Today's NLU?
Figure 4 for What's the Meaning of Superhuman Performance in Today's NLU?
Viaarxiv icon

We're Afraid Language Models Aren't Modeling Ambiguity

Apr 27, 2023
Alisa Liu, Zhaofeng Wu, Julian Michael, Alane Suhr, Peter West, Alexander Koller, Swabha Swayamdipta, Noah A. Smith, Yejin Choi

Figure 1 for We're Afraid Language Models Aren't Modeling Ambiguity
Figure 2 for We're Afraid Language Models Aren't Modeling Ambiguity
Figure 3 for We're Afraid Language Models Aren't Modeling Ambiguity
Figure 4 for We're Afraid Language Models Aren't Modeling Ambiguity
Viaarxiv icon