Retrieval pipelines, vector search, chunking and reranking: how models are grounded in a corpus instead of their weights.
arXiv:2608.03729v4 Announce Type: replace-cross Abstract: Automated Knowledge Base Construction (AKBC) is a core NLP task, and recent work proposes generating knowledge bases directly from large lang...
The paper introduces NAMOH, a native sparse attention mechanism that activates only a subset of heads per token, allowing each head to attend to a limited subsequence of tokens. By scaling the number of heads while keeping the active heads per token fixed, the method shortens head histories and reduces key‑value access without increasing overall storage. Experiments demonstrate that NAMOH can outperform fully activated models with the same parameter count and enable more efficient long‑context inference than smaller dense models.
BELIEFRAG is a closed‑loop controller that maintains an explicit evidence state—tracking sufficiency, reliability, conflict, uncertainty, evidence gaps, and acquisition cost—to guide adaptive retrieval‑augmented generation (RAG). By updating this state, it selects among retrieval, query rewriting, verification, answering, stopping, and abstention, achieving higher token‑F1 scores and lower token usage on six QA benchmarks compared to fixed iterative retrieval. The approach shows that corrective re‑retrieval and calibrated answerability are key to its performance gains.
Soft Spatial Reasoning introduces a post‑training framework for Large Vision‑Language Models that replaces hard, token‑by‑token chain‑of‑thought reasoning with a soft, continuous state formed by mixing token embeddings at each intermediate step. The method employs AdaptSoft, a controller that adjusts the degree of softness based on hidden states and predictive uncertainty, guided by a gradient‑alignment learning objective that requires no intermediate supervision. Experiments on diverse spatial benchmarks show that this approach outperforms both hard and fixed‑soft chain‑of‑thought baselines and several existing LVLMs.
The paper demonstrates that a stochastic hybrid system (SHS), which combines continuous dynamics governed by a stochastic differential equation (SDE) with discrete resets triggered by a Markov kernel, can be approximated by a single SDE in a higher‑dimensional latent space. By encoding the reset branches with auxiliary variables, the resets become deterministic, allowing the system’s manifold to be glued and embedded into Euclidean space. This embedding eliminates the need for explicit reset terms in the hybrid Fokker‑Planck equation, and the authors propose a loss function that matches evolving state distributions, enabling the latent SDE to recover the SHS’s probability evolution without mode labeling, trajectory segmentation, or event‑based simulations.
The paper introduces Actor‑Critic with Action Chunking (AC2), a method that assigns credit to short action chunks instead of entire trajectories, enabling policy updates without waiting for terminal rewards. AC2 employs local readiness, reference solutions, and 10k‑token chunks to make critic‑based credit assignment reliable. Experiments on Qwen3‑4B with FineProofs‑RL show AC2 surpasses GRPO’s peak validation score while using 2.5× fewer decoding FLOPs and fewer training steps.
The paper reports an empirical study comparing traditional and AI‑enabled Simulink controllers, using a taxonomy of ten structural categories and nine functional roles. Analyzing 62 real‑world models and surveying 13 practitioners, it finds that subsystem organization dominates all controller architectures, AI models rely heavily on discrete dynamics and user‑defined abstraction, and constraint enforcement blocks largely disappear in AI‑enabled designs. These findings highlight architectural tensions and gaps between AI literature and practical implementation.
The paper presents a controlled comparison of six retrieval-augmented generation (RAG) strategies for scientific question answering on a large arXiv corpus. All pipelines use the same LLM generator and evaluation protocol, differing only in retrieval design—ranging from classic dense retrieval to late‑interaction methods like ColBERTv2. The authors also release a synthetic question dataset and code to enable reproducible, large‑scale evaluation of RAG trade‑offs.
The paper treats multi‑path large language model reasoning as a diversity‑combining problem, analogous to noisy channel observations in wireless communications. It shows that the optimal symmetric linear combiner of latent embeddings is uniform, justifying majority vote in standard self‑consistency while allowing for weighting or pruning when prompt‑template branches are heterogeneous. By reducing path correlation through prompt‑template diversity, the authors propose an Adaptive‑K rule that selects an optimal number of reasoning paths, preserving most of the accuracy achieved with a fixed large number of paths across multiple models and benchmarks.
DualCast is a dual‑path language model that forecasts financial time‑series by combining a fast numerical forecaster with an optional text‑conditioned revision mechanism. The fast path trains only new financial‑token embeddings and output heads on a frozen Qwen3‑8B backbone, while the slow path uses a LoRA adapter to incorporate news and refine predictions. In zero‑shot tests across equities and energy prices at multiple time resolutions, the slow path achieves the lowest mean absolute percentage error in most settings, especially for longer horizons, and news ablations show additional gains in many markets.
The paper introduces an explainable pedestrian routing method that estimates perceived safety by first generating a natural‑language caption from street‑level images and then deriving risk scores solely from structured features of that caption. Benchmarking nine captioning setups against a CLIP image‑embedding baseline shows comparable performance, and the system was deployed on over 650,000 images across 36 wards in Manchester and Huddersfield. Field validation with 3,669 ratings from 70 participants revealed a modest but statistically significant correlation (r = 0.262) with human judgments, while a stronger supervised benchmark did not translate into better real‑world performance.
The paper introduces scTrilemma, a latent-bottleneck variational autoencoder designed to address the representation trilemma in single‑cell RNA‑seq data: preserving biological identity and state, remaining robust to nuisance context, and retaining gene‑level variation for expression analysis. scTrilemma routes expression‑derived variation to the embedding, decoder, or prior, gating gene tokens by expression and conditioning the prior on unlabeled pseudo‑bulk context, all under a single reconstruction objective without target annotations. In zero‑shot evaluations on successive CZ CELLxGENE Census releases, scTrilemma simultaneously satisfies all three demands, maintaining biological state, differential‑expression, and pathway structure across multiple disease settings, and latent interventions show context can be removed with minimal impact on other demands.
RAGScope is a leakage‑controlled, cost‑aware protocol that evaluates evidence‑gating mechanisms for retrieval‑augmented generation (RAG) systems using only the task input, retrieved context, and answer text. The enhanced gate, RAGScope‑E, achieves an AUROC of 0.798 and an average precision of 0.660 on three RAGTruth tasks, outperforming ROUGE‑L by 0.034 in pooled AP and delivering 0.748 precision within a top‑10% review budget. It operates quickly (6.22 ms per example on CPU) and demonstrates that cheap evidence gates can effectively triage RAG outputs, though calibration must be validated and adapted for each target domain.
The paper introduces STAR‑Ar, a BERT‑BiLSTM‑CRF model designed for the Daleel 2026 Arabic argument mining shared task. It treats argument discourse unit detection and classification as a token‑level sequence labeling problem, achieving an F1‑score of 72.69 on validation and 73.7 on test data. Analysis shows that models trained only on editorial texts perform worse than those trained on debates, mainly due to the smaller editorial dataset.
D‑Scope is a framework that links the interpretation of sparse autoencoder (SAE) features in diffusion transformers (DiTs) to controllable image generation. It aggregates SigLIP‑2 embeddings of highly activating image patches into visual centroids, matches target text descriptions against these centroids, and retrieves individual features without per‑feature text annotations. The method provides visual evidence for each selection and uses spatially masked interventions to test decoder directions under fixed generation conditions, evaluated across 150 SAEs and a benchmark of 100 target concepts.
TRACE is a lightweight learned selector that ranks completed search trajectories by aggregating cross‑rollout evidence, preserving individual query and evidence occurrences while propagating information across shared content or document identity. Trained with answer‑level supervision over frozen text embeddings, TRACE selects an existing answer without additional search or autoregressive aggregation, and a single selector generalizes across rollout policies and agent backbones. Across six WebQA policies, six long‑horizon dataset‑backbone combinations, and multiple WebQA benchmarks, TRACE outperforms majority voting and generative aggregators, achieving higher accuracy and at least tenfold higher processing throughput.
arXiv:2508.10148v2 Announce Type: replace-cross Abstract: Accurate and explainable out-of-distribution (OOD) detection is required to use machine learning systems safely. Previous work has shown that...
arXiv:2507.09601v3 Announce Type: replace-cross Abstract: Financial text embeddings must distinguish changes in event status, perspective, and obligations even when passages share similar wording. NM...
arXiv:2602.13194v3 Announce Type: replace-cross Abstract: Humans and large language models can predict next letter or word from its prior context much better than random guessing, indicating strong r...
arXiv:2607.13077v3 Announce Type: replace-cross Abstract: Large language models (LLMs) have been widely reported to homogenize human expression and thought. However, I argue that convergence in langu...