Retrieval pipelines, vector search, chunking and reranking: how models are grounded in a corpus instead of their weights.
arXiv:2609.24489v1 Announce Type: new Abstract: Offline reinforcement learning (RL) typically trains a critic by minimizing a regression loss against bootstrapped value targets stabilized by target n...
arXiv:2609.23769v1 Announce Type: new Abstract: The disambiguation of semantically similar statutory text across jurisdictions is a retrieval problem that existing methods do not solve. This inter-co...
arXiv:2609.23881v1 Announce Type: new Abstract: Joint Embedding Predictive Architectures (JEPAs) are a promising paradigm for learning task-agnostic latent world models without visual reconstruction....
arXiv:2609.24116v1 Announce Type: new Abstract: Although deep learning has advanced Whole Slide Image (WSI) Analysis, tissue artifacts like bubbles and folds often cause silent failures by concealing...
arXiv:2609.24136v1 Announce Type: new Abstract: Brain-to-image retrieval seeks to identify the visual stimulus that elicited a non-invasive neural response. Candidate images are typically represented...
arXiv:2609.24564v1 Announce Type: new Abstract: CLIP, a foundational vision-language model, has emerged as a powerful tool for open-vocabulary semantic segmentation. While freezing CLIP's text encode...
The paper introduces the Activity Chain Encoder (ACE), a self‑supervised deep learning model that transforms passively collected mobile phone location data into daily activity representations. ACE integrates pre‑trained urban embeddings, visit timing, and duration, using a Transformer to capture the sequential structure of stays, and is trained via masked activity modelling and contrastive learning without explicit activity labels. The resulting user‑level profiles are clustered and interpreted with temporal‑functional patterns and Census demographics, revealing six distinct weekday activity‑pattern groups in London that differ in daily rhythms, urban contexts, and demographic characteristics.
arXiv:2609. 22880v1 Announce Type: cross Abstract: LLM rerankers add of the order of \$0.
The paper introduces UniK, a universal knowledge perception platform designed to serve both digital AI—such as chatbots and agent workflows—and physical AI, which controls robots and autonomous systems. UniK handles the entire knowledge lifecycle—ingestion, enrichment, indexing, retrieval, and continuous evaluation—across diverse modalities including text, video, molecular data, and sensor telemetry, without task‑specific fine‑tuning. In five digital AI domains, UniK paired with a 70‑billion‑parameter model matches or surpasses larger proprietary LLMs, achieving high retrieval‑augmented generation accuracy on government data, medical QA, and chemistry tasks, and it also addresses similar data challenges in physical AI world‑model training.
The paper introduces Dial, a knowledge‑grounded framework for generating SQL queries that respect the specific dialect of heterogeneous database systems. Dial comprises a Dialect‑Aware Logical Query Planning module, a hierarchical intent‑aware knowledge base (HINT‑KB) containing canonical syntax, function, and constraint repositories, and an execution‑driven debugging loop that separates syntactic recovery from logic auditing. The authors also present DS‑NL2SQL, a benchmark of 2,218 dialect‑specific test cases across six major database systems, and report that Dial improves translation accuracy by 10.25% and dialect feature coverage by 15.77% over existing baselines.
The paper demonstrates that large language models (LLMs) used for forecasting real‑world events can be manipulated by simply publishing new articles, even without direct access to the model or its retriever. By injecting a small number of targeted news pieces into a common crawl corpus, an adversary can flip over half of the forecast probabilities and significantly degrade forecast accuracy. The study also shows that common defense strategies can be cheaply bypassed, highlighting the vulnerability of probabilistic LLM judgments to information‑supply‑chain attacks.
The paper introduces a pipeline and conversational system that processes 22,788 YouTube transcript and comment chunks from 309 North American cities to analyze public discourse on urbanism. It combines geographic entity resolution, topic modeling, sentiment analysis, and Retrieval-Augmented Generation (RAG), and reports empirical findings on model performance, such as a Twitter-tuned RoBERTa classifier outperforming VADER and dense retrieval surpassing TF‑IDF. The study also evaluates groundedness metrics, noting limitations of BERTScore and ROUGE‑1 for short user-generated text.
Re:CAP is a reference‑free audit loop for retrieval‑augmented generation (RAG) pipelines that probes for missing documents instead of enumerating all relevant ones. It identifies covered topics, generates probing questions, retrieves candidate documents, and uses an LLM judge to keep only those that add new information. On several benchmarks, Re:CAP recovers a significant portion of gold documents that flat BM25 or hybrid retrieval misses, and human evaluation shows most of these documents add new information.
The paper presents a smartphone‑compatible, retrieval‑augmented language model tailored to Bangladeshi statutory law. By distilling a 9‑billion‑parameter Gemma‑2 teacher into a 2‑billion‑parameter student using supervised fine‑tuning and QLoRA, the authors achieve significant gains in ROUGE‑L and BERTScore on an English benchmark while keeping the model lightweight (1.6 GB) and operable offline on a Pixel 6. The system retrieves from 36,029 statutory passages using a hybrid dense/BM25 approach, and cross‑lingual evaluation shows effective Bangla query handling against an English‑only corpus, with a practicing lawyer rating the responses highly in a pilot study.
The BD-LSC dataset introduces a bi‑directional lexical semantic change benchmark that tracks sense gain, loss, and stability across three time periods, while the ST‑WSD dataset offers fine‑grained, instance‑level sense annotations for words that blend slang and standard usage. These resources enable systematic evaluation of diverse models—including unsupervised clustering, supervised learning, transformer‑based approaches, and large language models—on tasks such as exact sense matching and multi‑label accuracy. The evaluation shows that few‑shot GPT‑4o performs best overall, yet all systems struggle with rare slang senses, highlighting a key open challenge in the field.
The paper presents a reproducible single‑subject baseline for reconstructing visual stimuli from EEG using a temporal‑spatial convolutional encoder that maps averaged EEG signals to 512‑dimensional ViT-B/32 image features. On the THINGS‑EEG2 dataset, the model achieves 12.83%, 39.17%, and 58.00% image recall at ranks 1, 5, and 10, respectively, outperforming analytical chance levels. The study also shows that performance drops sharply when applying a model trained on one subject to others, and that direct conditional generators without external visual weights produce noise‑dominated outputs, indicating that only coarse semantic decoding is feasible under the tested protocol.
Relationally Grounded Latent World Models for Autonomous Driving proposes using traffic scene graphs as privileged semantic supervision for latent world representations. The approach builds actor‑centric scene graphs from nuScenes 3D annotations, encodes their relational structure with a frozen text embedding model, and aligns visual latent representations to this semantic target during training. At inference the supervision branch is removed, requiring no scene graphs or 3D annotations and adding no extra computation, while achieving a 5.9% reduction in average trajectory L2 error and a 52.4% drop in collision rate compared to the LAW baseline.
The paper proposes a new sequence-to-sequence formulation for multi-view stereo (MVS) that jointly predicts 3D geometry for all input views using a global transformer architecture. It introduces ray‑map embeddings to inject camera parameters into image tokens and a unified global cost volume to capture 3D structure across all views. Experiments on public benchmarks demonstrate state‑of‑the‑art performance, outperforming both traditional MVS and feed‑forward reconstruction baselines.
SurgMotion is a video-native foundation model that replaces pixel-level reconstruction with latent motion prediction for surgical video analysis. It introduces motion-guided masked prediction, spatiotemporal affinity self-distillation, and spatiotemporal feature diversity regularization to focus on semantically meaningful regions and avoid representation collapse. Trained on SurgMotion-15M, the largest surgical video dataset, it outperforms state-of-the-art methods across 17 benchmarks, improving workflow recognition, action triplet recognition, skill assessment, polyp segmentation, and depth estimation.
The paper introduces the directional linear separability measure (D‑LSM) to quantify how much linear separability is preserved or improved by injective affine maps in neural networks. It characterizes the geometry supporting D‑LSM, proves its invariance under injective affine embeddings, and derives conditions for gated activations (ReLU, GELU, SiLU) to preserve and recover samples. Experiments validate the theoretical bounds, demonstrate affine‑tube constructions that achieve guaranteed recovery, and apply the method to Vision Transformer representations to obtain early post‑activation separability certificates.