AI
LEDGER: Claim-to-Evidence Trace Graphs for Auditing LLM Agents
Researchers have developed a system called LEDGER to help review and audit the outputs of large language models. The system creates layered graphs tha...
AI
AI
Vector Symbolic Policy Gradient
Researchers have developed a new method for training AI agents called Vector Symbolic Policy Gradient (VSPG). VSPG represents actions as unit-norm hyp...
AI
Mechanistic Interpretability of Structure-Aware Numerical Reasoning in LLaMA 3.1 8B
Researchers have investigated how the large language model Llama 3.1-8B performs in time-series prediction tasks. They created a task that requires un...
AI
Pedagogical AI in Mental Health: A Tri-Stream Fine-Tuned LLM Framework for Automated Clinical Supervision and Risk Triage
Researchers have developed a framework that uses artificial intelligence to help supervise mental health therapy sessions. The system analyzes audio a...
AI
Formal Verification of Romanov's Triplet Logic: A Verified Filter for Sliding-window 3-CNF with Application to Structured Formulas
Researchers have formally verified Romanov's Triplet Logic (TLS), a mathematical theory for reasoning about compatible paths through layered structure...
AI
ERASE: EaRly bAckpropagation SchEdule for Faster Training of Modern Recommendation Systems
Researchers propose ERASE, a scheduling technique for faster training of modern recommendation systems. Conventional training methods schedule forward...
AI
Coverage-Driven RTL Assertion Generation with Formal Exploration and Neuro-Symbolic Refinement
Researchers have developed a new system for generating high-quality assertions in hardware design. Assertions are like checks to ensure that digital d...
AI
Partition the Support, Reconstruct the Residual: Training-Free Sparse Attention for Video Generation and World Models
Researchers have developed a new method for accelerating video generation and world models using sparse attention. The approach, called SparsePR, invo...
AI
Physics-Unrolled Neural Operator for Wireless Field Modeling
Researchers have developed a new model called Physics-Unrolled Hybrid Neural Operator (PU-HNO) that predicts high-fidelity indoor radio maps from low-...
AI
Science Done on a Machine by a Machine: AI Agents in Computational Chemistry
Researchers are developing AI agents that can perform computational chemistry simulations autonomously, from designing experiments to analyzing result...
AI
OptiModNet: A UNet-Transformer Hybrid with Grouped-Query and Channel Attention for Optic Disc and Cup Segmentation
Researchers have proposed a new deep learning model called OptiModNet for segmenting the optic disc and cup in images. This task is critical for detec...
AI
GCNO: Gramian Chebyshev Neural Operator for Physics-Based Compression of Wireless Channels
Researchers have developed a new method for compressing wireless channel data called the Gramian Chebyshev Neural Operator (GCNO). This approach ident...