7 papers
The Residual Stream Is All You Need: On the Redundancy of the KV Cache in Transformer Inference
Kaleem Ullah Qasim, Jiashu Zhang, Muhammad Kafeel Shaheen +2
The key-value (KV) cache is widely treated as essential state in transformer inference, and a large body of work engineers policies to compress, evict, or approximate its entries.…
VERIFY-RL: Verifiable Recursive Decomposition for Reinforcement Learning in Mathematical Reasoning
Kaleem Ullah Qasim, Jiashu Zhang, Hao Li +1
Training language models to solve complex mathematical problems benefits from curriculum learning progressively training on simpler subproblems. However, existing decomposition met…
Accelerating Training Speed of Tiny Recursive Models with Curriculum Guided Adaptive Recursion
Kaleem Ullah Qasim, Jiashu Zhang
Background: Recursive reasoning models achieve strong performance through iterative refinement, allowing small networks to match large language models. However, training is computa…
ASTIF: Adaptive Semantic-Temporal Integration for Cryptocurrency Price Forecasting
Hafiz Saif Ur Rehman, Ling Liu, Kaleem Ullah Qasim
Financial time series forecasting is fundamentally an information fusion challenge, yet most existing models rely on static architectures that struggle to integrate heterogeneous k…
Complexity Agnostic Recursive Decomposition of Thoughts
Kaleem Ullah Qasim, Jiashu Zhang, Hafiz Saif Ur Rehman
Large language models often fail on multi-step reasoning due to fixed reasoning strategies that ignore problem specific difficulty. We introduce CARD (Complexity Agnostic Recursive…
MARBLE: A Multi-Agent Rule-Based LLM Reasoning Engine for Accident Severity Prediction
Kaleem Ullah Qasim, Jiashu Zhang
Accident severity prediction plays a critical role in transportation safety systems but is a persistently difficult task due to incomplete data, strong feature dependencies, and se…