Complex multi-hop question answering over text, tables, and images requires integrating heterogeneous evidence with inspectable reasoning. Single-LLM chain-of-thought is opaque, while planner–executor pipelines concentrate faults in a central plan, propagating early errors. We introduce DeALOG, a decentralized, central-planner-free multi-agent framework whose specialized agents, i.e., Table, Context, Visual, Summarizing, and Verification, coordinate only by reading and appending to a shared, typed natural-language log as persistent memory, with a lightweight scheduler rather than a planner. Rather than competing on raw accuracy, DeALOG targets what a central planner forfeits: robustness, since no single plan concentrates errors and evidence stays globally visible, and transparent, since the append-only log yields a provenance-anchored, human-readable reasoning trace. We position decentralized log-mediated coordination as a robustness-and-transparency primitive for multimodal reasoning

Diffusion language models generate text through iterative denoising, allowing intermediate predictions to influence later denoising steps, a recurring practice we call trajectory self-feedback (TSF). TSF is adopted under inconsistent terminology, evaluated on incompatible benchmarks, and studied by disconnected communities, yet no prior review characterises it reproducibly. We present a systematic literature review under an evidence-based methodology: six pre-specified questions, a documented six-library search with snowballing, explicit selection and quality criteria, and a structured synthesis of 24 primary studies across seven clusters, with a released audit trail. We find: (i) a model-agnostic taxonomy resolves four operator families: Prediction-Space Concatenation, Confidence-weighted Path Signals, Reward-Reinforced Conditioning, and Correction-Targeted Feedback, that the literature conflates as self-conditioning; (ii) evidence of benefit is real but fragmented across incomparable settings; (iii) within our search window, no study jointly varies the TSF mechanism and measures faithfulness on knowledge-intensive tasks; and (iv) a discrete/continuous asymmetry consistent with the sampling-wall hypothesis is visible but never tested. We conclude with eight open problems, a dependency map, and review-derived diagnostics (Commitment Index, Correction Capacity) and a Standardised Evaluation Protocol (SEP-DLM) to make TSF measurable across quality, correction, trajectory commitment, and faithfulness.

We present SENTINELEDGE, an open-source streaming detector for English-language scam phone calls that scores speech sentence-by- sentence and raises alerts while a conversa- tion is in progress, protecting callers rather than explaining losses afterward. It targets multi-minute scripted fraud calls, where sys- tems are closed and un-auditable and aca- demic alternatives use neural backbones un- suited to on-device budgets. SENTINELEDGE uses a streaming framework with per-sentence scoring, exponential moving average smooth- ing, and threshold-triggered alerts. It com- pares classifier heads using hand-crafted fea- tures, TF–IDF, SVM, gradient-boosted trees, and fine-tuned transformer on identical inputs and robustness conditions, revealing storage– accuracy trade-offs. We report bootstrap con- fidence intervals, stress-test ASR-style noise, adversarial scam phrasings and retraining, and zero-shot transfer to SMS and phishing URLs. Code, models, evaluation harness, and demon- stration are released under the MIT license

Recruiters and job seekers rely on search systems to navigate labor markets, making candidate matching engines critical for hiring outcomes. Most systems act as keyword filters, failing to handle skill synonyms and nonlinear careers, resulting in missed candidates and opaque match scores. We introduce JobMatchAI, a production-ready system integrating Transformer embeddings, skill knowledge graphs, and interpretable reranking. Our system optimizes utility across skill fit, experience, location, salary, and company preferences, providing factor-wise explanations through resume-driven search workflows. We release JobSearch-XS benchmark and a hybrid retrieval stack combining BM25, knowledge graph and semantic components to evaluate skill generalization. We assess system performance on JobSearch-XS across retrieval tasks, provide a demo video, a hosted website and installable package.

Federated Retrieval-Augmented Generation (Federated RAG) combines Federated Learning (FL),which enables distributed model training without exposing raw data, with Retrieval-Augmented Generation (RAG), which improves the factual accuracy of language models by grounding outputs in external knowledge. As large language models are increasingly deployed in privacy-sensitive domains such as healthcare, finance, and personalized assistance, Federated RAG offers a promising framework for secure, knowledge-intensive natural language processing (NLP). To the best of our knowledge, this paper presents the first systematic mapping study of Federated RAG, covering literature published between 2020 and 2025. Following Kitchenham’s guidelines for evidence-based software engineering, we develop a structured classification of research focuses, contribution types, and application domains. We analyze architectural patterns, temporal trends, and key challenges, including privacy-preserving retrieval, cross-client heterogeneity, and evaluation limitations. Our findings synthesize a rapidly evolving body of research, identify recurring design patterns, and surface open questions, providing a foundation for future work at the intersection of RAG and federated systems.