13 papers
To Call or Not to Call: A Framework to Assess and Optimize LLM Tool Calling
Qinyuan Wu, Soumi Das, Mahsa Amani +5
Agentic AI architectures augment LLMs with external tools, unlocking strong capabilities but potentially incurring substantial costs. Moreover, tool use is not always beneficial: r…
Revisiting the Past: Data Unlearning with Model State History
Keivan Rezaei, Mehrdad Saberi, Abhilasha Ravichander +1
Large language models are trained on massive corpora of web data, which may include private data, copyrighted material, factually inaccurate data, or data that degrades model perfo…
Fractional Rotation, Full Potential? Investigating Performance and Convergence of Partial RoPE
Mohammad Aflah Khan, Krishna P. Gummadi, Manish Gupta +1
Rotary Positional Embedding (RoPE) is a common choice in transformer architectures for encoding relative positional information. Although earlier work has examined omitting RoPE in…
In Agents We Trust, but Who Do Agents Trust? Latent Source Preferences Steer LLM Generations
Mohammad Aflah Khan, Mahsa Amani, Soumi Das +5
Agents based on Large Language Models (LLMs) are increasingly being deployed as interfaces to information on online platforms. These agents filter, prioritize, and synthesize infor…
What Has Been Lost with Synthetic Evaluation?
Alexander Gill, Abhilasha Ravichander, Ana MarasoviÄ
Large language models (LLMs) are increasingly used for data generation. However, creating evaluation benchmarks raises the bar for this emerging paradigm. Benchmarks must target sp…
The Curious Case of Factuality Finetuning: Models' Internal Beliefs Can Improve Factuality
Benjamin Newman, Abhilasha Ravichander, Jaehun Jung +5
Language models are prone to hallucination - generating text that is factually incorrect. Finetuning models on high-quality factual information can potentially reduce hallucination…