4 papers
Reward-Gated On-Policy Distillation
Mohammad Sadegh Akhondzadeh, Vijay Lingam, Atula Tejaswi +3
On-policy distillation is a powerful way to transfer reasoning ability from a strong teacher to a smaller student: the student samples trajectories from its own policy, and the tea…
One Sample is Enough to Make Conformal Prediction Robust
Soroush H. Zargarbashi, Mohammad Sadegh Akhondzadeh, Aleksandar Bojchevski
For any black-box model, conformal prediction (CP) returns prediction sets guaranteed to include the true label with high adjustable probability. Robust CP (RCP) extends the guaran…
EvA: Evolutionary Attacks on Graphs
Mohammad Sadegh Akhondzadeh, Soroush H. Zargarbashi, Jimin Cao +1
Even a slight perturbation in the graph structure can cause a significant drop in the accuracy of graph neural networks (GNNs). Most existing attacks leverage gradient information…
KurTail : Kurtosis-based LLM Quantization
Mohammad Sadegh Akhondzadeh, Aleksandar Bojchevski, Evangelos Eleftheriou +1
One of the challenges of quantizing a large language model (LLM) is the presence of outliers. Outliers often make uniform quantization schemes less effective, particularly in extre…