TechniqueRLHF / Alignment1 recent entries10 Sept 2026From Monolithic Blending to Agentic Orchestration: Dynamic Response for Conversational Assistants at ScalearXiv:2609.05758v2 Announce Type: new Abstract: Conversational assistants can blend retrieval, action selection, escalation, and wording in a single model path, or separate those roles. We report a pr
TechniqueRAG5 recent entries24 Sept 2024DoorDash deploys a Bedrock voice agent for Dasher supportDoorDash and AWS describe a live Bedrock-backed voice agent for Dasher support, reporting hundreds of thousands of calls daily.
TechniqueAgents8 recent entries8 Jan 2026Zapier reports company-wide AI adoption and support workflow resultsZapier reports 97% daily-work AI use and says its AI support-ticket summarizer halved average handle time.→17 Apr 2026Don't Retrieve, Navigate: Distilling Enterprise Knowledge into Navigable Agent Skills for QA and RAGarXiv:2604.14572v1 Announce Type: cross Abstract: Retrieval-Augmented Generation (RAG) grounds LLM responses in external evidence but treats the model as a passive consumer of search results: it never→1 Jun 2026The Architecture of Errors: From Universal Impossibility to Patch-Local LLM ReliabilityarXiv:2605.30628v1 Announce Type: cross Abstract: Universal LLM reliability is not a finite-library problem: across all possible tasks, tools, schemas, knowledge sources, and evaluator expectations, n→24 Aug 2026Knowledge-Graph-Gated Defactualization for Style-Controllable and Fact-Preserving Generation in Agentic Conversational AIarXiv:2608.20393v1 Announce Type: cross Abstract: Agentic large language models (LLMs) deployed in fact-sensitive applications such as customer support must simultaneously preserve factual correctness→26 Aug 2026AgentWorld: Personality-Aware Reliability Evaluation for Agentic Information RetrievalarXiv:2608.24076v1 Announce Type: new Abstract: Evaluation of agentic information retrieval remains limited to scripted interactions with uniform users, missing both natural personality diversity and →10 Sept 2026From Monolithic Blending to Agentic Orchestration: Dynamic Response for Conversational Assistants at ScalearXiv:2609.05758v2 Announce Type: new Abstract: Conversational assistants can blend retrieval, action selection, escalation, and wording in a single model path, or separate those roles. We report a pr→21 Sept 2026When Better Turns Do Not Make Better Agents: Diagnosing the Gap Between Next-Turn Metrics and Workflow SuccessarXiv:2609.21187v1 Announce Type: new Abstract: Agent models are frequently evaluated one decision at a time, where the model predicts the next action based on the gold interaction history, which is s→9 Oct 2026RIT-RAG: Navigating Document Corpora with Retrieval-Induced TreesarXiv:2610.11370v1 Announce Type: new Abstract: Retrieval-augmented generation (RAG) grounds language models in external corpora. Agentic RAG enables iterative search, yet exposes the model to isolate
TechniqueFine-tuning4 recent entries24 Aug 2026Knowledge-Graph-Gated Defactualization for Style-Controllable and Fact-Preserving Generation in Agentic Conversational AIarXiv:2608.20393v1 Announce Type: cross Abstract: Agentic large language models (LLMs) deployed in fact-sensitive applications such as customer support must simultaneously preserve factual correctness→26 Aug 2026AgentWorld: Personality-Aware Reliability Evaluation for Agentic Information RetrievalarXiv:2608.24076v1 Announce Type: new Abstract: Evaluation of agentic information retrieval remains limited to scripted interactions with uniform users, missing both natural personality diversity and →21 Sept 2026When Better Turns Do Not Make Better Agents: Diagnosing the Gap Between Next-Turn Metrics and Workflow SuccessarXiv:2609.21187v1 Announce Type: new Abstract: Agent models are frequently evaluated one decision at a time, where the model predicts the next action based on the gold interaction history, which is s→24 Sept 2026Can One Adapted Model Do It All? Fine-Tuning Strategy Selection for Customer Support LLMsarXiv:2609.27262v1 Announce Type: new Abstract: Production customer-support systems often require LLMs to support multiple skills, such as intent classification, question answering, summarization, or
TechniqueSafety1 recent entries10 Sept 2026From Monolithic Blending to Agentic Orchestration: Dynamic Response for Conversational Assistants at ScalearXiv:2609.05758v2 Announce Type: new Abstract: Conversational assistants can blend retrieval, action selection, escalation, and wording in a single model path, or separate those roles. We report a pr