N. Udaya Shankar

Mohaddeseh Bastan, Nishant Shankar, Mihai Surdeanu et al.

Can language models read biomedical texts and explain the biomedical mechanisms discussed? In this work we introduce a biomedical mechanism summarization task. Biomedical studies often investigate the mechanisms behind how one entity (e.g., a protein or a chemical) affects another in a biological context. The abstracts of these publications often include a focused set of sentences that present relevant supporting statements regarding such relationships, associated experimental evidence, and a concluding sentence that summarizes the mechanism underlying the relationship. We leverage this structure and create a summarization task, where the input is a collection of sentences and the main entities in an abstract, and the output includes the relationship and a sentence that summarizes the mechanism. Using a small amount of manually labeled mechanism sentences, we train a mechanism sentence classifier to filter a large biomedical abstract collection and create a summarization dataset with 22k instances. We also introduce conclusion sentence generation as a pretraining task with 611k instances. We benchmark the performance of large bio-domain language models. We find that while the pretraining task help improves performance, the best model produces acceptable mechanism outputs in only 32% of the instances, which shows the task presents significant challenges in biomedical language understanding and summarization.

Maheep Chaudhary, Ian Su, Nikhil Hooda et al.

Large language models (LLMs) can internally distinguish between evaluation and deployment contexts, a behaviour known as \emph{evaluation awareness}. This undermines AI safety evaluations, as models may conceal dangerous capabilities during testing. Prior work demonstrated this in a single $70$B model, but the scaling relationship across model sizes remains unknown. We investigate evaluation awareness across $15$ models scaling from $0.27$B to $70$B parameters from four families using linear probing on steering vector activations. Our results reveal a clear power-law scaling: evaluation awareness increases predictably with model size. This scaling law enables forecasting deceptive behavior in future larger models and guides the design of scale-aware evaluation strategies for AI safety. A link to the implementation of this paper can be found at https://anonymous.4open.science/r/evaluation-awareness-scaling-laws/README.md.

Kaushal Bhogale, Nishant Shankar, Adheesh Juvekar et al.

Although modern face verification systems are accessible and accurate, they are not always robust to pose variance and occlusions. Moreover, accurate models require a large amount of data to train. We structure our experiments to operate on small amounts of data obtained from an NGO that funds ophthalmic surgeries. We set up our face verification task as that of verifying pre-operation and post-operation images of a patient that undergoes ophthalmic surgery, and as such the post-operation images have occlusions like an eye patch. In this paper, we present a system that performs the face verification task using one-shot learning. To this end, our paper uses deep convolutional networks and compares different model architectures and loss functions. Our best model achieves 85% test accuracy. During inference time, we also attempt to detect image forgeries in addition to performing face verification. To achieve this, we use Error Level Analysis. Finally, we propose an inference pipeline that demonstrates how these techniques can be used to implement an automated face verification and forgery detection system.