Retrieval pipelines, vector search, chunking and reranking: how models are grounded in a corpus instead of their weights.
arXiv:2505.13994v3 Announce Type: replace Abstract: Retrieval-Augmented Generation (RAG) systems empower large language models (LLMs) with external knowledge, yet struggle with efficiency-accuracy tr...
arXiv:2605.16638v2 Announce Type: replace Abstract: Recent research has demonstrated that Universal Multimodal Embedding (UME) benefits significantly from Chain-of-Thought (CoT) reasoning. In this pa...
arXiv:2609.25143v1 Announce Type: new Abstract: Unsupervised Continual chunking is a fundamental problem in machine learning and neuroscience, where the goal is to identify groups of states that freq...
arXiv:2609.25541v1 Announce Type: new Abstract: Tabular foundation models learn to predict cell values in context, whereas world-model self-supervision asks for prediction in representation space (Le...
arXiv:2609.26217v1 Announce Type: new Abstract: Compact language models are typically deployed by retaining a single post-training checkpoint and sampling it repeatedly. In this work, we challenge th...
arXiv:2609.25411v1 Announce Type: cross Abstract: Classifier-free Guidance (CFG) is widely adopted in text-to-speech (TTS) systems to enhance generation quality and conditioning fidelity by interpola...
arXiv:2605.24437v2 Announce Type: replace Abstract: We present a novel framework for embedding hard constraint satisfaction into neural network (NN) architectures, specifically feedforward neural net...
arXiv:2503.04332v2 Announce Type: replace-cross Abstract: The tremendous commercial potential of large language models (LLMs) has heightened concerns over their unauthorized use. To address this, we...
arXiv:2609.25537v1 Announce Type: new Abstract: Large language model (LLM) inference is constrained by the quadratic scaling of self-attention and the linear scaling of the KV cache, increasing laten...
arXiv:2609.26218v1 Announce Type: cross Abstract: Structural graph analysis of the academic publishing network captures the topological relationships between entities but does not see the content of...
arXiv:2606.22807v3 Announce Type: replace Abstract: As retrieval systems scale, effective and efficient reranking becomes increasingly important. However, most existing encoder- and decoder-based rer...
arXiv:2609.26073v1 Announce Type: new Abstract: Pathology foundation models (PFMs) provide strong tile-level representations but remain difficult to interpret at the cellular and microenvironmental s...
arXiv:2609.26443v1 Announce Type: new Abstract: Recent advances in Artificial Intelligence (AI)-powered Computer-Aided Diagnosis (CAD) systems have substantially improved breast cancer screening, dia...
LADDER is a new framework that combines diffusion language modeling with Graph Retrieval-Augmented Generation (GraphRAG) to enable efficient multi‑hop reasoning. It introduces an event‑driven self‑clocking retrieval mechanism that triggers graph queries only when new entities appear, and an incomplete‑query graph propagation module that aggregates multi‑hop evidence during parallel decoding. Experiments on three multi‑hop QA benchmarks show that LADDER improves exact match from 39.6% to 45.2% while reducing latency by 4.1×.
DUMA-Bench is a new benchmark that evaluates the security of large language model agents in dual‑control settings, where both the agent and the user can modify the shared environment. It builds on the existing τ²‑bench by adding adversarial environments that cover eight vulnerability classes, such as RAG poisoning and unsafe output handling. The authors tested 14 models from five families and found that dual‑control interaction raises attack success rates from 26.9% to 41.1%, demonstrating that agent security depends on the interaction between model, user, and environment.
SemDHT introduces a certified semantic index for discovering agent-accessible capabilities over exact-key distributed hash tables (DHTs). It uses a two-layer semantic sketch—coarse cells for grouping descriptors and residual codes for refining candidate selection—allowing providers to publish at a bounded set of derived keys while requesters probe precision keys first. Anchor committees certify descriptor-to-key consistency, enabling efficient, consistent discovery with fewer lookups and reduced publication fan‑out, as demonstrated by high recall and significant speedups over locality‑sensitive hashing in real‑world experiments.
The paper introduces a physics‑guided deep metric learning approach for open‑world radar pulse de‑interleaving, leveraging continuous Time‑of‑Arrival sinusoidal positional encodings to model physical inter‑pulse durations. It builds on a transformer‑based framework, optimizing network parameters solely with Supervised Contrastive learning and employing physics‑based priors—PRI consistency and AoA continuity—for validation and checkpoint selection via unsupervised HDBSCAN clustering. This method aims to improve de‑interleaving performance in dense, contested electromagnetic environments where classical techniques falter.
The paper introduces ufakzeka-1, a 151‑million‑parameter Turkish language model trained from scratch on 13.5 B tokens. It details the tokenizer, a three‑stage pretraining schedule, post‑training data augmentation, and a comprehensive evaluation suite that includes release gates, a rule‑checked conversation sweep, and hand tests—all with prompts excluded from training data. The authors report three key findings about safety gate performance, training‑seed variance, and data‑round effects, and they release the model weights, data recipe, evaluation code, and spend ledger under Apache‑2.0.
The paper introduces modality‑gated deep adapters, a parameter‑efficient method for adding new modalities to a frozen multimodal embedding language model without altering its existing outputs. These adapters are bottleneck modules attached to each decoder layer, grouped into modality‑specific packs that activate only during encoding of their own modality, ensuring exact preservation of the base model’s computation graph. Experiments on a 2B base model show significant gains in audio‑to‑text and thermal‑to‑text retrieval metrics, and the authors release the audio and thermal packs along with training and evaluation code.
Orthrus is a serving system that performs heterogeneous batching of embedding and generative models within a single inference loop. It uses chunked embedding with incremental pooling and workload‑aware batch composition to unify conflicting computational patterns. Experiments on four A100 GPUs show that Orthrus improves throughput by 1.28×–4.52× and reduces p99 latency by up to 55.8% compared to baseline deployments.