803 citations · 2.6k across the 256 of their papers we have counts for
15 papers · 1 filter
The Role of Feed-Forward Layers in Transformer Dynamics
Thomas Jacob Maranzatto, Semih Akkoc, Sennur Ulukus
We study the dynamical behavior of tokens in transformers from a control-theoretic perspective. Our model includes the feed-forward layer present after the self-attention mechanism…
On the Diverse Dynamical Behaviors Arising in Deep Linear Transformers
Sixu Li, Thomas Jacob Maranzatto, Jan Peszek +5
We study the inference-time behavior of deep linear encoder-only transformers through the lens of interacting particle systems. In this perspective, tokens are modeled as particles…
RunAgent: Interpreting Natural-Language Plans with Constraint-Guided Execution
Arunabh Srivastava, Mohammad A., Khojastepour +2
Humans solve problems by executing targeted plans, yet large language models (LLMs) remain unreliable for structured workflow execution. We propose RunAgent, a multi-agent plan exe…
Multi-Modal Semantic Communication
Matin Mortaheb, Erciyes Karakaya, Sennur Ulukus
Semantic communication aims to transmit information most relevant to a task rather than raw data, offering significant gains in communication efficiency for applications such as te…
Re-ranking the Context for Multimodal Retrieval Augmented Generation
Matin Mortaheb, Mohammad A. Amir Khojastepour, Srimat T. Chakradhar +1
Retrieval-augmented generation (RAG) enhances large language models (LLMs) by incorporating external knowledge to generate a response within a context with improved accuracy and re…
RAG-Check: Evaluating Multimodal Retrieval Augmented Generation Performance
Matin Mortaheb, Mohammad A. Amir Khojastepour, Srimat T. Chakradhar +1
Retrieval-augmented generation (RAG) improves large language models (LLMs) by using external knowledge to guide response generation, reducing hallucinations. However, RAG, particul…