163 citations · 430 across the 19 of their papers we have counts for
21 papers
ML-AutoResearch: Training Machine Learning Research Agents with Automatically Generated Environments
Ziyang Cai, Amir Saeidi, Harkirat Behl
With the advent of AI agents, automated scientific discovery is becoming an increasingly plausible goal. However, training agents to autonomously execute the engineering-heavy labo…
The Bias is in the Details: An Assessment of Cognitive Bias in LLMs
R. Alexander Knipper, Charles S. Knipper, Kaiqi Zhang +3
As Large Language Models (LLMs) are increasingly embedded in real-world decision-making processes, it becomes crucial to examine the extent to which they exhibit cognitive biases.…
Sample More to Think Less: Group Filtered Policy Optimization for Concise Reasoning
Vaishnavi Shrivastava, Ahmed Awadallah, Vidhisha Balachandran +3
Large language models trained with reinforcement learning with verifiable rewards tend to trade accuracy for length--inflating response lengths to achieve gains in accuracy. While…
Jailbreak Distillation: Renewable Safety Benchmarking
Jingyu Zhang, Ahmed Elgohary, Xiawei Wang +5
Large language models (LLMs) are rapidly deployed in critical applications, raising urgent needs for robust safety benchmarking. We propose Jailbreak Distillation (JBDistill), a no…
Phi-4-reasoning Technical Report
Marah Abdin, Sahaj Agarwal, Ahmed Awadallah +20
We introduce Phi-4-reasoning, a 14-billion parameter reasoning model that achieves strong performance on complex reasoning tasks. Trained via supervised fine-tuning of Phi-4 on car…
Phi-4 Technical Report
Marah Abdin, Jyoti Aneja, Harkirat Behl +24
We present phi-4, a 14-billion parameter language model developed with a training recipe that is centrally focused on data quality. Unlike most language models, where pre-training…
Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone
Marah Abdin, Jyoti Aneja, Hany Awadalla +126
We introduce phi-3-mini, a 3.8 billion parameter language model trained on 3.3 trillion tokens, whose overall performance, as measured by both academic benchmarks and internal test…
PEEKABOO: Interactive Video Generation via Masked-Diffusion
Yash Jain, Anshul Nasery, Vibhav Vineet +1
Modern video generation models like Sora have achieved remarkable success in producing high-quality videos. However, a significant limitation is their inability to offer interactiv…
Unlocking Spatial Comprehension in Text-to-Image Diffusion Models
Mohammad Mahdi Derakhshani, Menglin Xia, Harkirat Behl +2
We propose CompFuser, an image generation pipeline that enhances spatial comprehension and attribute assignment in text-to-image generative models. Our pipeline enables the interpr…
DAMEX: Dataset-aware Mixture-of-Experts for visual understanding of mixture-of-datasets
Yash Jain, Harkirat Behl, Zsolt Kira +1
Construction of a universal detector poses a crucial question: How can we most effectively train a model on a large mixture of datasets? The answer lies in learning dataset-specifi…
Efficiently Robustify Pre-trained Models
Nishant Jain, Harkirat Behl, Yogesh Singh Rawat +1
A recent trend in deep learning algorithms has been towards training large scale models, having high parameter count and trained on big dataset. However, robustness of such large s…
Textbooks Are All You Need
Suriya Gunasekar, Yi Zhang, Jyoti Aneja +16
We introduce phi-1, a new large language model for code, with significantly smaller size than competing models: phi-1 is a Transformer-based model with 1.3B parameters, trained for…
Generalized Decoding for Pixel, Image, and Language
Xueyan Zou, Zi-Yi Dou, Jianwei Yang +11
We present X-Decoder, a generalized decoding model that can predict pixel-level segmentation and language tokens seamlessly. X-Decodert takes as input two types of queries: (i) gen…
Neural-Sim: Learning to Generate Training Data with NeRF
Yunhao Ge, Harkirat Behl, Jiashu Xu +6
Training computer vision models usually requires collecting and labeling vast amounts of imagery under a diverse set of scene configurations and properties. This process is incredi…
Scaling the Convex Barrier with Sparse Dual Algorithms
Alessandro De Palma, Harkirat Singh Behl, Rudy Bunel +2
Tight and efficient neural network bounding is crucial to the scaling of neural network verification systems. Many efficient bounding algorithms have been presented recently, but t…
AutoSimulate: (Quickly) Learning Synthetic Data Generation
Harkirat Singh Behl, Atılım Güneş Baydin, Ran Gal +2
Simulation is increasingly being used for generating large labelled datasets in many machine learning problems. Recent methods have focused on adjusting simulator parameters with t…
STEER: Simple Temporal Regularization For Neural ODEs
Arnab Ghosh, Harkirat Singh Behl, Emilien Dupont +2
Training Neural Ordinary Differential Equations (ODEs) is often computationally expensive. Indeed, computing the forward pass of such models involves solving an ODE which can becom…
Progressive Skeletonization: Trimming more fat from a network at initialization
Pau de Jorge, Amartya Sanyal, Harkirat S. Behl +3
Recent studies have shown that skeletonization (pruning parameters) of networks \textit{at initialization} provides all the practical benefits of sparsity both at inference and tra…
Alpha MAML: Adaptive Model-Agnostic Meta-Learning
Harkirat Singh Behl, Atılım Güneş Baydin, Philip H. S. Torr
Model-agnostic meta-learning (MAML) is a meta-learning technique to train a model on a multitude of learning tasks in a way that primes the model for few-shot learning of new tasks…
Meta Learning Deep Visual Words for Fast Video Object Segmentation
Harkirat Singh Behl, Mohammad Najafi, Anurag Arnab +1
Personal robots and driverless cars need to be able to operate in novel environments and thus quickly and efficiently learn to recognise new object classes. We address this problem…
Incremental Tube Construction for Human Action Detection
Harkirat Singh Behl, Michael Sapienza, Gurkirt Singh +3
Current state-of-the-art action detection systems are tailored for offline batch-processing applications. However, for online applications like human-robot interaction, current sys…