Picture for Robert Nowak

Robert Nowak

Task Vectors in In-Context Learning: Emergence, Formation, and Benefit

Add code
Jan 16, 2025
Viaarxiv icon

Deep Active Learning in the Open World

Add code
Nov 10, 2024
Viaarxiv icon

AHA: Human-Assisted Out-of-Distribution Generalization and Detection

Add code
Oct 10, 2024
Figure 1 for AHA: Human-Assisted Out-of-Distribution Generalization and Detection
Figure 2 for AHA: Human-Assisted Out-of-Distribution Generalization and Detection
Figure 3 for AHA: Human-Assisted Out-of-Distribution Generalization and Detection
Figure 4 for AHA: Human-Assisted Out-of-Distribution Generalization and Detection
Viaarxiv icon

SIEVE: General Purpose Data Filtering System Matching GPT-4o Accuracy at 1% the Cost

Add code
Oct 03, 2024
Figure 1 for SIEVE: General Purpose Data Filtering System Matching GPT-4o Accuracy at 1% the Cost
Figure 2 for SIEVE: General Purpose Data Filtering System Matching GPT-4o Accuracy at 1% the Cost
Figure 3 for SIEVE: General Purpose Data Filtering System Matching GPT-4o Accuracy at 1% the Cost
Figure 4 for SIEVE: General Purpose Data Filtering System Matching GPT-4o Accuracy at 1% the Cost
Viaarxiv icon

Humor in AI: Massive Scale Crowd-Sourced Preferences and Benchmarks for Cartoon Captioning

Add code
Jun 15, 2024
Figure 1 for Humor in AI: Massive Scale Crowd-Sourced Preferences and Benchmarks for Cartoon Captioning
Figure 2 for Humor in AI: Massive Scale Crowd-Sourced Preferences and Benchmarks for Cartoon Captioning
Figure 3 for Humor in AI: Massive Scale Crowd-Sourced Preferences and Benchmarks for Cartoon Captioning
Figure 4 for Humor in AI: Massive Scale Crowd-Sourced Preferences and Benchmarks for Cartoon Captioning
Viaarxiv icon

Pretraining Decision Transformers with Reward Prediction for In-Context Multi-task Structured Bandit Learning

Add code
Jun 07, 2024
Figure 1 for Pretraining Decision Transformers with Reward Prediction for In-Context Multi-task Structured Bandit Learning
Figure 2 for Pretraining Decision Transformers with Reward Prediction for In-Context Multi-task Structured Bandit Learning
Figure 3 for Pretraining Decision Transformers with Reward Prediction for In-Context Multi-task Structured Bandit Learning
Figure 4 for Pretraining Decision Transformers with Reward Prediction for In-Context Multi-task Structured Bandit Learning
Viaarxiv icon

SaVeR: Optimal Data Collection Strategy for Safe Policy Evaluation in Tabular MDP

Add code
Jun 04, 2024
Figure 1 for SaVeR: Optimal Data Collection Strategy for Safe Policy Evaluation in Tabular MDP
Figure 2 for SaVeR: Optimal Data Collection Strategy for Safe Policy Evaluation in Tabular MDP
Figure 3 for SaVeR: Optimal Data Collection Strategy for Safe Policy Evaluation in Tabular MDP
Figure 4 for SaVeR: Optimal Data Collection Strategy for Safe Policy Evaluation in Tabular MDP
Viaarxiv icon

Future Prediction Can be a Strong Evidence of Good History Representation in Partially Observable Environments

Add code
Feb 11, 2024
Figure 1 for Future Prediction Can be a Strong Evidence of Good History Representation in Partially Observable Environments
Figure 2 for Future Prediction Can be a Strong Evidence of Good History Representation in Partially Observable Environments
Figure 3 for Future Prediction Can be a Strong Evidence of Good History Representation in Partially Observable Environments
Figure 4 for Future Prediction Can be a Strong Evidence of Good History Representation in Partially Observable Environments
Viaarxiv icon

Learning from the Best: Active Learning for Wireless Communications

Add code
Jan 23, 2024
Figure 1 for Learning from the Best: Active Learning for Wireless Communications
Figure 2 for Learning from the Best: Active Learning for Wireless Communications
Figure 3 for Learning from the Best: Active Learning for Wireless Communications
Figure 4 for Learning from the Best: Active Learning for Wireless Communications
Viaarxiv icon

DIRECT: Deep Active Learning under Imbalance and Label Noise

Add code
Dec 14, 2023
Figure 1 for DIRECT: Deep Active Learning under Imbalance and Label Noise
Figure 2 for DIRECT: Deep Active Learning under Imbalance and Label Noise
Figure 3 for DIRECT: Deep Active Learning under Imbalance and Label Noise
Viaarxiv icon