arXiv:2609.20981v1 Announce Type: new Abstract: Autoregressive (AR) models suffer from local greediness, while diffusion language models (DLMs) often lack…
Tag: cs.AI updates on arXiv.org
Detecting Hallucination in LLMs: Tracing the Topological Signatures of Impaired Context Sharing
arXiv:2609.21096v1 Announce Type: new Abstract: In this work, we examine the topology of information flow patterns within attention graphs to effectively…
Attention-Aware Routing: Coupling Routing and Attention in MoEs
arXiv:2609.20974v1 Announce Type: new Abstract: In Mixture-of-Experts language models, the router typically selects and weights experts based on the…
LoRA Enhanced Contrastive Learning with SAS Vision Transformers
arXiv:2609.21061v1 Announce Type: new Abstract: Automatic target recognition (ATR) with synthetic aperture sonar (SAS) supports advanced naval…
Generating a Consistent Enterprise: Synthesis and Reference-Free Evaluation of Multi-System Business Data
arXiv:2609.11286v2 Announce Type: replace Abstract: Synthetic relational data is normally produced by a model trained on a real dataset, and its quality…
Atria Dawn: The Dawn of Agentic Superintelligence
arXiv:2609.15818v2 Announce Type: replace Abstract: As AI agents become participants in the development of their successors, they reshape both the…
The Other Half of the Memory Wall: Serving 35B MoEs from SSD with Trained Routing Prediction
arXiv:2609.18063v2 Announce Type: replace Abstract: Mixture-of-experts (MoE) inference on consumer hardware is bounded by weight memory: a 35B-class model…
AquiLLM: Evaluating Faithfulness in Open-Weight RAG-LLM Systems for Scientific Research
arXiv:2609.16519v2 Announce Type: replace Abstract: Scientific research increasingly relies on large, heterogeneous data sources, motivating interest in…
Inference-Time Nash Alignment
arXiv:2609.08082v2 Announce Type: replace Abstract: Preference-based fine-tuning methods such as RLHF and DPO require substantial compute and large…
Attributing Preprocessing Invariance in Spectral Foundation Models
arXiv:2608.14227v3 Announce Type: replace Abstract: A spectral foundation model should remain useful when laboratories preprocess spectra differently. The…