Retrieval pipelines, vector search, chunking and reranking: how models are grounded in a corpus instead of their weights.
arXiv:2608.28776v1 Announce Type: new Abstract: Multilingual sentence embeddings are increasingly used to estimate semantic similarity across languages, yet their sensitivity to fine-grained translat...
arXiv:2608.29702v1 Announce Type: new Abstract: A token-embedding table holds a hub of short rows near its origin, and we show that this cluster biases what nearest-neighbor intrinsic-dimension (ID)...
arXiv:2608.31007v1 Announce Type: cross Abstract: Understanding how psychiatric patients subjectively experienced a clinical conversation is important for feedback and alliance-related process monito...
arXiv:2605.30947v4 Announce Type: replace Abstract: LLM-based research agents have advanced rapidly in science and engineering, where research is organized around executable experiments, code, and qu...
arXiv:2608.28663v1 Announce Type: new Abstract: Reliable individual cattle identification supports disease surveillance, vaccination records, breeding management, and livestock insurance. Although th...
arXiv:2608.31065v1 Announce Type: new Abstract: Surgical phase recognition is key to context-aware computer-assisted feedback in vitreoretinal procedures, yet the scarcity of synchronized multimodal...
arXiv:2604.22280v4 Announce Type: replace Abstract: Multimodal Large Language Models (MLLMs) have emerged as a promising foundation for universal multimodal embeddings. Recent studies have shown that...
The paper introduces a GPU‑optimized retrieval framework for LinkedIn’s semantic search, partitioning embeddings into eight category‑supervised segments and applying a min/median aggregation rule aligned with the existing relevance policy. A lightweight Stage‑1 scorer generates high‑recall candidates, while a two‑stage GPU architecture—FP8 coarse ranking followed by FP16 re‑ranking—boosts throughput and recall, achieving 99.6‑99.8% of full‑FP16 recall at over 500 QPS per shard. In A/B testing, the system raises exploratory‑query Precision@10 from 63.7% to 79.0% and navigational Precision@1 from 65.5% to 74.7%, with human evaluation confirming the improvement.
The paper introduces CERES, a closed‑loop multimodal indexing framework that addresses semantic collapse in multimodal generation by building a three‑level semantic pyramid and using scale‑routed cross‑attention to generate images that remain retrievable by their original queries. CERES employs a co‑occurrence‑aware router, a lightweight U‑Net generator, and a soft‑Jaccard coverage objective to ensure generated images cover the intended concepts, verified by re‑indexing with a frozen vision‑language model and an external DINOv2 probe. Experiments on four pansharpening benchmarks show state‑of‑the‑art performance, especially under extreme scale variation, and significant improvements in concept‑query retrieval and image‑text ranking metrics.
BEACON is a tri‑modal contrastive learning framework that enriches AlphaEarth embeddings by aligning physical representations from Earth‑observation imagery with semantic POI text and human behavioral POI visitation data, while keeping the deployed model image‑only. In a Houston case study, BEACON outperformed six baselines on nine downstream tasks, achieving up to 43% higher R² for obesity prevalence, 34% for poor mental health, and 22% for median household income under a linear probe.
arXiv:2608.29001v1 Announce Type: new Abstract: In a data-driven world, efficiently organizing and mapping relationships between objects is crucial. Graphs are powerful tools for modeling these conne...
AOI-Net introduces a structural face AOI-guided Eye‑Gaze Track Network that jointly models short‑term temporal dynamics and AOI‑level structural organization for Autism Spectrum Disorder detection. The network uses a gating mechanism to adaptively combine complementary representations and incorporates class‑distribution‑aware learning to address the imbalance between ASD and typically developing participants. Experiments on a large clinical eye‑tracking database with over 1,300 participants demonstrate that AOI‑Net outperforms state‑of‑the‑art methods and offers interpretable gaze‑behavior modeling for scalable AI‑driven ASD screening.
The paper investigates when text embeddings can be reliably used as inputs for empirical analysis by formalizing the assumption that embeddings preserve essential information from the original text. Under a generative model where documents are mixtures of latent topics, the author examines two applications: clustering documents in embedding space and using embeddings as controls for high‑dimensional text. Empirical results on 363 U.S. metropolitan areas show that clusters derived from LLM‑generated economic descriptions capture interpretable economic archetypes and better distinguish local employment dynamics than alternative clustering methods or curated covariates.
The paper investigates why token prediction, a common pre‑training objective for language models, yields useful representations. It introduces a statistical framework linking token prediction accuracy to the geometry of token embeddings, showing that accurate predictions organize embeddings according to Hellinger distances between context distributions. The authors also propose a self‑consistency principle that refines contextual representations through repeated application of a shared block, and provide downstream guarantees for token generation, community recovery, and linear classification.
AgenticRag‑R1 is a reinforcement‑learning framework that integrates reasoning, retrieval, and memory through a stack and fine‑grained action space. It uses hierarchical action‑aware rewards and an information‑aware trajectory rejection strategy to support long‑horizon learning. Experiments on multi‑hop, open‑domain, and agentic reasoning benchmarks show that AgenticRag‑R1 outperforms strong baselines and produces robust, interpretable, memory‑aware reasoning behaviors.
Flow-JEPA introduces a conditional flow matching dynamics model that generates a sequence of future latent states conditioned on current observations and actions, replacing deterministic autoregressive prediction with stochastic trajectory-level prediction. By using a Gaussian flow source, the model learns to transport perturbed latent trajectories toward clean future representations while remaining within the reconstruction‑free JEPA framework. The approach improves mean success rates from 86% to 92% under clean observations and from 67% to 86% under noisy conditions.
The paper presents the Temporal Coherence Score (TCS), a continuous, interpretable metric for detecting temporal inconsistencies in political news. TCS is computed through a four‑stage pipeline that extracts temporal facts, builds a temporal knowledge graph, verifies consistency using internal rules and external references, and aggregates scores with explanations. On a benchmark of 100 political articles with injected errors, TCS achieves 0.909 precision, providing detailed explanations for each flagged inconsistency.
arXiv:2606.13061v3 Announce Type: replace Abstract: Reasoning-driven universal multimodal embedding has advanced rapidly by introducing Chain-of-Thought (CoT) reasoning into the embedding pipeline. D...
arXiv:2608.30316v1 Announce Type: new Abstract: Existing class-incremental learning methods struggle in multi-label scenarios (MLCIL) due to the inherent contradiction of learning objectives arising...
SpIDER is a dense retrieval method that combines LLM reasoning with graph-based exploration of codebases to locate relevant functions, classes, or files for user queries. It introduces a graph-structured benchmark, SpIDER-Bench, covering multiple programming languages and demonstrates significant recall improvements over traditional BM25 and dense approaches. The method’s graph-based candidate expansion provides auditable structural reasons for each retrieved item while keeping the retrieval budget fixed.