6 papers
Countdown-Code: A Testbed for Studying The Emergence and Generalization of Reward Hacking in RLVR
Muhammad Khalifa, Zohaib Khan, Omer Tafveez +2
Reward hacking is a form of misalignment in which models overoptimize proxy rewards without genuinely solving the underlying task. Precisely measuring reward hacking occurrence rem…
To Lie or Not to Lie? Investigating The Biased Spread of Global Lies by LLMs
Zohaib Khan, Mustafa Dogan, Ifeoma Okoh +6
Misinformation is on the rise, and the strong writing capabilities of LLMs lower the barrier for malicious actors to produce and disseminate false information. We study how LLMs be…
Plasticity vs. Rigidity: The Impact of Low-Rank Adapters on Reasoning on a Micro-Budget
Zohaib Khan, Omer Tafveez, Zoha Hayat Bhatti
Recent advances in mathematical reasoning typically rely on massive scale, yet the question remains: can strong reasoning capabilities be induced in small language models ($\leq1.5…
Scaling Truth: The Confidence Paradox in AI Fact-Checking
Ihsan A. Qazi, Zohaib Khan, Abdullah Ghani +7
The rise of misinformation underscores the need for scalable and reliable fact-checking solutions. Large language models (LLMs) hold promise in automating fact verification, yet th…
With a Grain of SALT: Are LLMs Fair Across Social Dimensions?
Samee Arif, Zohaib Khan, Maaidah Kaleem +3
This paper presents a systematic analysis of biases in open-source Large Language Models (LLMs), across gender, religion, and race. Our study evaluates bias in smaller-scale Llama…
Beyond Uniform Query Distribution: Key-Driven Grouped Query Attention
Zohaib Khan, Muhammad Khaquan, Omer Tafveez +2
The Transformer architecture has revolutionized deep learning through its Self-Attention mechanism, which effectively captures contextual information. However, the memory footprint…