24 citations · 109 across the 19 of their papers we have counts for
24 papers
LLM Agents Can Easily Tamper With Their Own Traces
Jeremy Qin, David Schmotz, Derck Prinzhorn +3
Asynchronous monitoring, incident investigations, and compliance audits primarily rely on agent traces to reconstruct what happened. These analyses assume that LLM agents cannot ta…
Instrumental Monitor Evasion Emerges Under Ordinary Task Pressure
David Schmotz, Derck Prinzhorn, Luca Beurer-Kellner +3
A central concern in AI safety is that agents may treat oversight as an obstacle when it conflicts with completing their goals. We study instrumental evasion, the propensity of LLM…
Open Problems in Machine Unlearning for AI Safety
Fazl Barez, Tingchen Fu, Ameya Prabhu +16
As AI systems become more capable, widely deployed, and increasingly autonomous in critical areas such as cybersecurity, biological research, and healthcare, ensuring their safety…
Wu's Method can Boost Symbolic AI to Rival Silver Medalists and AlphaGeometry to Outperform Gold Medalists at IMO Geometry
Shiven Sinha, Ameya Prabhu, Ponnurangam Kumaraguru +2
Proving geometric theorems constitutes a hallmark of visual reasoning combining both intuitive and logical skills. Therefore, automated theorem proving of Olympiad-level geometry p…
No "Zero-Shot" Without Exponential Data: Pretraining Concept Frequency Determines Multimodal Model Performance
Vishaal Udandarao, Ameya Prabhu, Adhiraj Ghosh +5
Web-crawled pretraining datasets underlie the impressive "zero-shot" evaluation performance of multimodal models, such as CLIP for classification/retrieval and Stable-Diffusion for…
Efficient Lifelong Model Evaluation in an Era of Rapid Progress
Ameya Prabhu, Vishaal Udandarao, Philip Torr +3
Standardized benchmarks drive progress in machine learning. However, with repeated testing, the risk of overfitting grows as algorithms over-exploit benchmark idiosyncrasies. In ou…
Corrective Machine Unlearning
Shashwat Goel, Ameya Prabhu, Philip Torr +2
Machine Learning models increasingly face data integrity challenges due to the use of large-scale training datasets drawn from the Internet. We study what model developers can do i…
Random Representations Outperform Online Continually Learned Representations
Ameya Prabhu, Shiven Sinha, Ponnurangam Kumaraguru +3
Continual learning has primarily focused on the issue of catastrophic forgetting and the associated stability-plasticity tradeoffs. However, little attention has been paid to the e…
From Categories to Classifiers: Name-Only Continual Learning by Exploring the Web
Ameya Prabhu, Hasan Abed Al Kader Hammoud, Ser-Nam Lim +3
Continual Learning (CL) often relies on the availability of extensive annotated datasets, an assumption that is unrealistically time-consuming and costly in practice. We explore a…
Inverse Scaling: When Bigger Isn't Better
Ian R. McKenzie, Alexander Lyzhov, Michael Pieler +24
Work on scaling laws has found that large language models (LMs) show predictable improvements to overall loss with increased scale (model size, training data, and compute). Here, w…
Rapid Adaptation in Online Continual Learning: Are We Evaluating It Right?
Hasan Abed Al Kader Hammoud, Ameya Prabhu, Ser-Nam Lim +3
We revisit the common practice of evaluating adaptation of Online Continual Learning (OCL) algorithms through the metric of online accuracy, which measures the accuracy of the mode…
Online Continual Learning Without the Storage Constraint
Ameya Prabhu, Zhipeng Cai, Puneet Dokania +3
Traditional online continual learning (OCL) research has primarily focused on mitigating catastrophic forgetting with fixed and limited storage allocation throughout an agent's lif…
Real-Time Evaluation in Online Continual Learning: A New Hope
Yasir Ghunaim, Adel Bibi, Kumail Alhamoud +5
Current evaluations of Continual Learning (CL) methods typically assume that there is no constraint on training time and computation. This is an unrealistic assumption for any real…
Computationally Budgeted Continual Learning: What Does Matter?
Ameya Prabhu, Hasan Abed Al Kader Hammoud, Puneet Dokania +4
Continual Learning (CL) aims to sequentially train models on streams of incoming data that vary in distribution by preserving previous knowledge while adapting to new data. Current…
Towards Adversarial Evaluations for Inexact Machine Unlearning
Shashwat Goel, Ameya Prabhu, Amartya Sanyal +3
Machine Learning models face increased concerns regarding the storage of personal user data and adverse impacts of corrupted data like backdoors or systematic bias. Machine Unlearn…
No Cost Likelihood Manipulation at Test Time for Making Better Mistakes in Deep Networks
Shyamgopal Karthik, Ameya Prabhu, Puneet K. Dokania +1
There has been increasing interest in building deep hierarchy-aware classifiers that aim to quantify and reduce the severity of mistakes, and not just reduce the number of errors.…
Simple Unsupervised Multi-Object Tracking
Shyamgopal Karthik, Ameya Prabhu, Vineet Gandhi
Multi-object tracking has seen a lot of progress recently, albeit with substantial annotation costs for developing better and larger labeled datasets. In this work, we remove the n…
"You might also like this model": Data Driven Approach for Recommending Deep Learning Models for Unknown Image Datasets
Ameya Prabhu, Riddhiman Dasgupta, Anush Sankaran +2
For an unknown (new) classification dataset, choosing an appropriate deep learning architecture is often a recursive, time-taking, and laborious process. In this research, we propo…
Sampling Bias in Deep Active Classification: An Empirical Study
Ameya Prabhu, Charles Dognin, Maneesh Singh
The exploding cost and time needed for data labeling and model training are bottlenecks for training DNN models on large datasets. Identifying smaller representative data samples w…
Hybrid Binary Networks: Optimizing for Accuracy, Efficiency and Memory
Ameya Prabhu, Vishal Batchu, Rohit Gajawada +2
Binarization is an extreme network compression approach that provides large computational speedups along with energy and memory savings, albeit at significant accuracy costs. We in…
Distribution-Aware Binarization of Neural Networks for Sketch Recognition
Ameya Prabhu, Vishal Batchu, Sri Aurobindo Munagala +2
Deep neural networks are highly effective at a range of computational tasks. However, they tend to be computationally expensive, especially in vision-related problems, and also hav…
Deep Expander Networks: Efficient Deep Networks from Graph Theory
Ameya Prabhu, Girish Varma, Anoop Namboodiri
Efficient CNN designs like ResNets and DenseNet were proposed to improve accuracy vs efficiency trade-offs. They essentially increased the connectivity, allowing efficient informat…
Towards Deep Learning in Hindi NER: An approach to tackle the Labelled Data Scarcity
Vinayak Athavale, Shreenivas Bharadwaj, Monik Pamecha +2
In this paper we describe an end to end Neural Model for Named Entity Recognition NER) which is based on Bi-Directional RNN-LSTM. Almost all NER systems for Hindi use Language Spec…
Towards Sub-Word Level Compositions for Sentiment Analysis of Hindi-English Code Mixed Text
Ameya Prabhu, Aditya Joshi, Manish Shrivastava +1
Sentiment analysis (SA) using code-mixed data from social media has several applications in opinion mining ranging from customer satisfaction to social campaign analysis in multili…