Scaling Search-Augmented Reasoning Agents via Adaptive Information Control
Siheng Xiong, Oguzhan Gungordu, Blair Johnson, James C. Kerce, Faramarz Fekri

Project

DeepControl is an adaptive framework that optimizes information retrieval and expansion based on an agent’s reasoning state, enhancing performance across various benchmarks.

Planning through World Model for Automated Heuristic Design via Self-Evolving LLMs
Oguzhan Gungordu, Siheng Xiong, Faramarz Fekri

Project

PathWise is a multi-agent framework that uses stateful memory and evolutionary actions to guide automated heuristic design, enabling structured reasoning, reuse of prior derivations, and controlled self-evolution of LLM-generated heuristics.

Enhancing Language Model Reasoning with Structured Multi-Level Modeling
Siheng Xiong, Ali Payani, Faramarz Fekri

Project

Multi-Level Reasoning (MLR) is a lightweight planner-executor loop that improves long-horizon reasoning, using iterative Step-DPO for scalable supervision to enhance accuracy and stability.

Deliberate Reasoning in Language Models as Structure-Aware Planning with an Accurate World Model
Siheng Xiong, Ali Payani, Yuan Yang, Faramarz Fekri

Project

SWAP (Structure-Aware Planning) is a framework for multi-step reasoning with LMs, where the world model predicts the next state as a graph, guiding the policy model to propose the next action.

CausalEval: Towards Better Causal Reasoning in Language Models
Longxuan Yu*, Delin Chen*, Siheng Xiong*, Qingyang Wu, Dawei Li, Zhikai Chen, Xiaoze Liu, Liangming Pan

Project

We provide a comprehensive review of research aimed at enhancing LMs for causal reasoning. We evaluate the performance of different LMs and methods on various causal reasoning tasks, providing key findings and in-depth analysis.

Can LLMs Reason in the Wild with Programs?
Yuan Yang, Siheng Xiong, Ali Payani, Ehsan Shareghi, Faramarz Fekri

Project

Tiger is a TactIc-Guided ReasonER designed to tackle reasoning-in-the-wild tasks by generating and refining programs. It learns from previous trajectories to iteratively improve program generation, enabling more effective reasoning (like OpenAI o1).

Large Language Models Can Learn Temporal Reasoning
Siheng Xiong, Ali Payani, Ramana Kompella, Faramarz Fekri

Project

TG-LLM performs temporal reasoning in two steps: 1) Text-to-Temporal Graph translation: generate temporal graph given the context and keyword; 2) Temporal Graph Reasoning: perform deliberate CoT reasoning over the temporal graph.

Harnessing the power of large language models for natural language to first-order logic translation
Yuan Yang, Siheng Xiong, Ali Payani, Ehsan Shareghi, Faramarz Fekri

Project

LogicLLaMA can be used standalone or to correct previously generated rules by other models for the NL-FOL translation task.

Long-Context Modeling with Dynamic Hierarchical Sparse Attention for On-Device LLMs
Siheng Xiong, Joe Zou, Faramarz Fekri, Yae Jee Cho

Project

We introduce Dynamic Hierarchical Sparse Attention (DHSA), a plug-in module for Transformers that improves efficiency by predicting token-level sparsity using chunk-level similarity, reducing latency and memory usage while maintaining performance.

The Compressor-Retriever Architecture for Language Model OS
Yuan Yang, Siheng Xiong, Ehsan Shareghi, Faramarz Fekri

Project

We introduce compressor-retriever, a model-agnostic architecture designed for life-long context management. Our approach exclusively uses the base model’s forward function to compress and retrieve context, ensuring end-to-end differentiability.

TILR: Temporal Inductive Logic Reasoning over Hypergraphs
Yuan Yang, Siheng Xiong, Ali Payani, James C Kerce, Faramarz Fekri

Project

TILR is a reasoning framework that detects inductive patterns in temporal data via neural-logic methodology. The framework aims to assist the training of modern ML models by inducing patterns for accurate grounding with fewer data.

TEILP: Time prediction over knowledge graphs via logical reasoning
Siheng Xiong, Yuan Yang, Ali Payani, James C Kerce, Faramarz Fekri

Project

TEILP extends TILP by converting TKGs into temporal event knowledge graphs (TEKGs) and developing a differentiable random walk approach with conditional probability density functions for time prediction based on query intervals.

TILP: Differentiable learning of temporal logical rules on knowledge graphs
Siheng Xiong, Yuan Yang, Faramarz Fekri, James Clayton Kerce

Project

TILP is the first differentiable framework for learning temporal logical rules, using a constrained random walk mechanism and temporal operators to model key temporal features.