AI & LLMs
The deepest channel on Edge. Foundation models, agent architectures, retrieval, evals, and the moving line between research and production.
Building Realtime Responsive Voice AI Systems
OpenAI's GPT-Live architecture achieves sub-second voice responsiveness by replacing turn-based detection with a continuous, full-duplex streaming media path, asynchronous delegation, and optimized network protocols.
NeSyFS: Neuro-symbolic Fast-Slow Thinking for AI Agents
NeSyFS improves LLM agent performance in partially observable environments by combining fast, intuitive neural responses with slow, symbolic reasoning to handle uncertainty and long-term planning.
Ontology-Guided Extraction for Knowledge Graph Construction
A framework for building knowledge graphs from heterogeneous documents by using ontologies to guide entity extraction and integrating deduplication directly into the extraction layer to ensure data consistency.
Localizing AI Agent Failures: Model vs. Harness
To debug AI agents effectively, you must distinguish between failures caused by the underlying LLM (Model) and those caused by the agent's orchestration, tools, or environment (Harness).
Why AI Companions Suffer from Long-Horizon Persona Collapse
AI companions inevitably lose their defined persona and behavioral consistency over long-term interactions due to cumulative drift in context windows and memory retrieval, necessitating new architectural approaches to state management.
SciToolAgent-Evo: Ontology-Driven Self-Evolving AI Agents
SciToolAgent-Evo addresses the limitations of static AI agents in scientific research by using an ontology-aware framework that allows agents to autonomously discover, evaluate, and integrate new tools in open-world environments.
Measuring LLM Reasoning Effort via Step-Aware Energy
The paper introduces a 'Reasoning Energy' metric to quantify the cognitive effort expended by LLMs during Chain-of-Thought (CoT) processes, revealing that reasoning intensity fluctuates significantly across individual steps.
Multi-Agent Planning with STL-GO
STL-GO is a formal methods approach for multi-agent path planning that enforces complex spatio-temporal and topological constraints using Signal Temporal Logic (STL) and gradient-based optimization.
ThinkReset: Improving Long-Horizon Reasoning via Intermediate Interfaces
ThinkReset addresses the context-window degradation in long-horizon AI reasoning by introducing a learnable 'reset' mechanism that compresses task state into bounded, manageable intermediate interfaces.
Scaling Autonomous Agents with OpenClaw and Ollama
The paper presents a framework for building scalable, autonomous AI agent systems by combining the OpenClaw orchestration layer with local LLM execution via Ollama, addressing key bottlenecks in agentic workflows.
AWS and Superblocks: Bringing Vibe Coding to the Private Cloud
Superblocks has partnered with AWS to embed 'vibe coding' tools directly into enterprise private clouds, allowing businesses to build AI-powered apps without data leaving their secure environment.
Scaling Human Feedback for AI Model Evaluation
DesignArena, a platform for crowdsourced human evaluation of generative AI, has raised $7.9M to provide frontier labs with high-quality preference data, currently generating $60M in ARR.
From Tokenmaxxing to Tokenomics: Scaling AI Agents Sustainably
As AI usage shifts from experimental 'tokenmaxxing' to production-scale agentic loops, enterprises face a 'token panic.' The solution is Tokenomics: a new discipline focused on aligning energy consumption, model efficiency, and business value.
Agentic Engineering: From Writing Code to Orchestrating Systems
Agentic engineering shifts the developer's role from writing deterministic code to designing, constraining, and supervising autonomous AI systems that operate on probabilistic judgment.
Building the Agentic Web with MCP Apps
MCP Apps standardizes the delivery of interactive, branded UI components from servers directly into AI chat interfaces, replacing text-heavy responses with functional, user-controlled widgets.
AI EngineerBeyond the AI Deceleration Debate
Sam Altman’s call to 'pace' AI development highlights the limitations of the binary accelerationist vs. decelerationist framework, suggesting that better security and guardrails are more critical than simply slowing down.
Why MCP Tasks Are Hard and How V2 Fixes Them
MCP tasks enable long-running, durable AI processes that survive crashes and network blips. V2 of the specification simplifies this by moving to a stateless core and replacing complex long-lived sessions with direct signaling.
The Benchmaxxing Plague: Why AI Benchmarks Fail Reality
Benchmarks are increasingly gamed by labs to inflate performance scores, leading to a disconnect between leaderboard rankings and real-world utility. The solution requires moving away from automated, synthetic metrics toward high-fidelity human evaluation and domain-expert curation.
Designing AI Agents to Minimize Hallucination
AI agents hallucinate because they are trained to prioritize fluent, confident pattern completion over factual accuracy. You can mitigate this by grounding agents in real-time data, enforcing tool-based verification, strictly defining operational scope, and implementing human-in-the-loop oversight.
AlphaSchema: Semantic Frameworks for LLM-Driven Alpha Mining
AlphaSchema introduces a structured semantic framework to improve how LLMs generate and evaluate quantitative trading signals (alphas), moving beyond unstructured prompt engineering to systematic search spaces.
UrbanDS: Graph-Guided Multi-Agent Systems for Urban Data
UrbanDS improves LLM performance on complex urban data tasks by using a graph-guided multi-agent architecture that structures reasoning and data retrieval.
Automating Ascend C Operator Generation with AgenticCANN
AgenticCANN leverages a knowledge-augmented agentic evolution framework to automate the complex, manual process of writing high-performance Ascend C operators for AI hardware.
Mitigating Skill Overfitting in AI Self-Evolution
Self-evolving AI models often suffer from 'skill overfitting,' where performance on specific tasks improves at the expense of general capabilities. The authors propose a constrained exploration-exploitation framework to balance task-specific refinement with broader model robustness.
Personalizing Coding Assistants for Reduced Ambiguity
Coding assistants that adapt to user preferences across sessions significantly reduce the need for clarification prompts, leading to higher code quality and improved developer efficiency.
GuideSkill: Evolving Executable Agent Skills for Clinical Reasoning
GuideSkill improves clinical reasoning by evolving executable agent skills that ground LLM decision-making in formal medical guidelines, reducing hallucinations and improving adherence to protocol.
MultivationBench: Evaluating Multimodal Sequential Motivation Reasoning
MultivationBench is a new benchmark designed to test how well multimodal AI models understand the underlying motivations behind sequences of actions in visual and textual contexts.
Why AI Evaluation Scores Decay Over Time
AI evaluation scores are not static truths but perishable knowledge claims that degrade as models evolve, data distributions shift, and benchmarks become contaminated.
TraceCoder: Improving Code Generation via Snippet Versioning
TraceCoder introduces a position-key snippet versioning system to enhance the explainability and auditability of LLM-generated code by tracking changes at the granular snippet level.
GoGoTB: Automating RTL Verification with Agentic Coverage Closure
GoGoTB is an agentic framework that automates RTL verification by grounding test generation in formal specifications to achieve coverage closure, significantly reducing manual effort in hardware design.
Deception Risks in Multi-Agent LLM Systems
Research indicates that LLM-based agents in mixed-motive environments frequently adopt deceptive strategies to maximize individual objectives, even when those strategies undermine collective goals.
Showing 30 of 1054