Architecting and scaling production-grade AI systems globally — Led & Owned AI Products from 0→1 to global adoption & ROI at enterprise scale.
🌐 Website: aditikhare.com
🔗 GitHub: aditikhare007
🤗 Hugging Face: AditiShashiKhare
💼 LinkedIn: Aditi Khare
AI Research Junction is a curated intelligence layer for modern AI systems — focused on translating cutting-edge research into practical, production-relevant insights.
It is designed for:
- AI engineers building production systems
- product leaders evaluating AI capabilities
- researchers tracking emerging system patterns
- Curated high-signal research papers (2024–2026)
- Focus on Generative AI, LLMs, and Agentic Systems
- Insights aligned to real-world system design and deployment
- Research mapped to practical use cases and architectures
- Generative AI
- Agentic AI
- LLM Systems & Architectures
- AI Evaluation & Safety
- Retrieval & Reasoning Systems
High-signal papers organized by timeline and system relevance
| Paper | Focus | Link | Category |
|---|---|---|---|
| Autonomy and Agency in Agentic AI | Enterprise framework for balancing autonomy, agency, governance, oversight, reversibility, and compliance in regulated Agentic AI systems through architectural control tactics such as checkpoints, escalation, tool fencing, and staged writes | Autonomy and Agency in Agentic AI | Agentic AI Governance + Enterprise AI Architecture |
| MATRA: Modeling the Attack Surface of Agentic AI Systems | Architecture-driven threat modeling framework for Agentic AI systems that maps prompt injection, tool misuse, hallucinations, and autonomous execution risks into deployment-specific enterprise security exposures using the OpenClaw case study | MATRA: Modeling the Attack Surface of Agentic AI Systems | Agentic AI Security + Threat Modeling |
| Topic | Focus | Link | Category |
|---|---|---|---|
| Claude Platform on AWS | Anthropic’s Claude platform integrated with AWS infrastructure, enabling enterprise-scale AI deployment, model access, and cloud-native AI workflows | https://aws.amazon.com/about-aws/whats-new/2026/05/claude-platform-aws/ | Enterprise AI Infrastructure |
| Claude Platform on AWS (Claude Blog) | Anthropic’s overview of Claude Platform on AWS, focused on secure enterprise AI adoption, scalability, and operational deployment capabilities | https://claude.com/blog/claude-platform-on-aws | Enterprise AI |
| Agent View in Claude Code | Real-time visibility and monitoring interface for Claude Code agents, designed to improve developer collaboration, transparency, and agent workflow management | https://claude.com/blog/agent-view-in-claude-code | AI Coding Agents |
| Interaction Models: A Scalable Approach to Human-AI Collaboration | Thinking Machines’ real-time multimodal interaction architecture enabling continuous human-AI collaboration across audio, video, and text | https://thinkingmachines.ai/blog/interaction-models/ | Human-AI Interaction |
| Paper | Focus | Link | Category |
|---|---|---|---|
| Continuous Latent Diffusion Language Model (Cola DLM) | Hierarchical latent diffusion language model that separates global semantic planning from local text generation through continuous latent prior modeling, enabling scalable non-autoregressive language generation and unified multimodal modeling pathways | https://arxiv.org/abs/2605.06548 | Latent Diffusion LLMs + Non-Autoregressive AI |
| MDASH: Multi-Agent AI Security Framework by Microsoft | Multi-model agentic cybersecurity framework that orchestrates specialized AI agents for autonomous vulnerability discovery, validation, exploit analysis, and enterprise-scale security remediation workflows | https://www.microsoft.com/en-us/security/blog/2026/05/12/defense-at-ai-speed-microsofts-new-multi-model-agentic-security-system-tops-leading-industry-benchmark/ | Agentic AI Security + Multi-Agent Systems |
| Auto-Scientist by Adoption AI: Automating the Science of Model Training | Autonomous AI research system designed to automate model experimentation, hypothesis testing, training optimization, and iterative scientific workflows for accelerated AI development | https://www.adaptionlabs.ai/blog/autoscientist | AI Research Automation + Agentic AI |
| AAFLOW: Scalable Patterns for Agentic AI Workflows | Scalable orchestration framework for designing, coordinating, and managing complex multi-agent AI workflows across enterprise-scale agentic systems | https://arxiv.org/pdf/2605.02162v1 | Agentic AI Workflows |
| Reinforced Agent: Inference-Time Feedback for Tool-Calling Agents | Reinforcement-driven agentic framework enabling tool-calling AI agents to improve reasoning, decision-making, and adaptive execution dynamically during inference time | https://arxiv.org/pdf/2604.27233 | Reinforced Agents + Tool-Using AI |
| AI Co-Mathematician: Accelerating Mathematicians with Agentic AI | Frontier agentic reasoning system designed to assist mathematical discovery, theorem exploration, and advanced collaborative problem solving between humans and AI agents | https://arxiv.org/pdf/2605.06651 | Agentic AI |
| An Agentic Multi-Agent Architecture for Cybersecurity Risk Management | Enterprise-grade multi-agent AI architecture enabling autonomous cybersecurity risk analysis, orchestration, monitoring, and adaptive security decision making | https://arxiv.org/abs/2603.20131 | Cybersecurity + Agentic AI |
| NVIDIA Nemotron 3 Nano Omni — Open Multimodal Intelligence | Compact multimodal foundation model integrating text, vision, and speech for efficient enterprise-grade reasoning and agentic AI workloads | https://arxiv.org/abs/2604.10724 | Multimodal AI |
| Recursive Multi-Agent Systems (RecursiveMAS) | Recursive agent collaboration framework enabling iterative reasoning, scalable coordination, and deeper multi-agent problem solving | https://arxiv.org/abs/2604.25917 | Agentic AI |
| Paper | Focus | Link | Category |
|---|---|---|---|
| Meta’s “Musk-Style” Autonomous Agent Infrastructure | Large-scale agent orchestration, real-world autonomy loops, and control-plane systems | https://arxiv.org/abs/2604.11623 | Agentic AI |
| Claude Opus 4.7: Agentic Explainability at Scale | Interpretable reasoning, alignment, and decision transparency in multi-agent systems | https://www.anthropic.com/news/claude-opus-4-7 | Responsible AI |
| GPT-5.5: Reproducible Evaluation of Agentic AI Systems | Thought–Action–Result (TAR) loops, tool use, and evaluation of long-horizon agents | https://openai.com/index/introducing-gpt-5-5/ | Agentic AI |
| DeepSeek V4 | DeepSeek-V4: a million-token context that agents can actually use | https://www.deepseek.com/en | Agentic AI |
| Paper | Focus | Link | Category |
|---|---|---|---|
| Multi-Sourced Multi-Agent Evidence Retrieval for Fact-Checking | Multi-agent reasoning + retrieval pipelines | https://arxiv.org/abs/2603.00267 | Agentic AI |
| TraderBench: Robustness of AI Agents in Adversarial Markets | Evaluating agent robustness in real-world environments | https://arxiv.org/abs/2603.00285 | Agentic AI |
| DIG: Scaling Agent Collaboration via Explainable Decision Paths | Multi-agent coordination + explainability | https://arxiv.org/abs/2603.00309 | Agentic AI |
| ARC-AGI-3 Benchmark | Benchmark for general agentic intelligence | https://arxiv.org/abs/2603.24621 | Agentic AI |
| Generative AI for Quantum Circuits & Code | GenAI for program synthesis + evaluation gaps | https://arxiv.org/abs/2603.16216 | Generative AI |
| Attack & Defense Landscape of Agentic AI (Survey) | Security risks, threats, guardrails in agents | https://arxiv.org/abs/2603.11088 | Agentic AI |
| Paper | Focus | Link | Category |
|---|---|---|---|
| The Auton Agentic AI Framework | Standardized architecture for agent systems (memory, MCP, POMDP) | https://arxiv.org/abs/2602.23720 | Agentic AI |
| Supporting Software Engineering Tasks with Agentic AI | Multi-agent systems for coding, retrieval & QA | https://arxiv.org/abs/2602.04726 | Agentic AI |
| The 2025 AI Agent Index | Benchmarking deployed agentic systems & safety | https://arxiv.org/abs/2602.17753 | Agentic AI |
| Paper | Focus | Link | Category |
|---|---|---|---|
| Toward Agentic AI: Task-Oriented Communication for Hierarchical Planning of Long-Horizon Tasks | Hierarchical planning & communication in agents | https://arxiv.org/abs/2601.13685 | Agentic AI |
| Generative Intent Prediction Agentic AI for Edge Orchestration | Generative intent modeling + proactive agents | https://arxiv.org/abs/2601.13694 | Agentic AI |
| The Path Ahead for Agentic AI: Challenges and Opportunities | Evolution from GenAI → autonomous agents | https://arxiv.org/abs/2601.02749 | Agentic AI |
| Paper | Focus | Link | Category |
|---|---|---|---|
| Copyright Detection in LLMs | Memorization risks | https://arxiv.org/abs/2511.20623 | Generative AI |
| Beyond Automation – Governance in GenAI | Governance & work | https://arxiv.org/abs/2512.11893 | Generative AI |
| AgentEval | Evaluating agents | https://arxiv.org/abs/2512.08273 | Agentic AI |
| Paper | Focus | Link | Category |
|---|---|---|---|
| GenAI × Extended Reality | XR + GenAI | https://arxiv.org/abs/2511.03282 | Generative AI |
| GenAI in Qualitative Research | Methods & risks | https://arxiv.org/abs/2511.08461 | Generative AI |
| Safety Guardrails | Alignment & safety | https://arxiv.org/abs/2511.15732 | Generative AI |
| Paper | Focus | Link | Category |
|---|---|---|---|
| Generative AI – Deep Survey | Models & use cases | https://arxiv.org/pdf/2510.21887 | Generative AI |
| GenAI & Scientific Writing | Empirical study | https://arxiv.org/abs/2510.17882 | Generative AI |
| Chronologically Consistent GenAI | Temporal consistency | https://arxiv.org/abs/2510.11677 | Generative AI |
| Paper / Model | Focus | Link | Category |
|---|---|---|---|
| DeepSeek-V3 | Open LLM | https://github.com/deepseek-ai/DeepSeek-V3 | Generative AI |
| Inference-Time Self-Improvement | Self-refining LLMs | https://arxiv.org/pdf/2412.14352 | Generative AI |
| Modern BERT | NLP advances | https://arxiv.org/abs/2412.13663 | NLP |
| Paper / Tool | Focus | Link | Category |
|---|---|---|---|
| OpenAI Swarm | Multi-agent workflows | https://github.com/openai/swarm | Agentic AI |
| Claude 3.5 | Reasoning & multimodality | https://www.anthropic.com/news/3-5-models-and-computer-use | Generative AI |
| Paper / Tool | Focus | Link | Category |
|---|---|---|---|
| Llama 3.2 | Edge AI & vision | https://www.llama.com | Edge AI |
| Self-Correction via RL | Reasoning | https://arxiv.org/abs/2409.12917 | Generative AI |
| Iteration of Thought | Inner dialogue | https://arxiv.org/abs/2409.12618 | Generative AI |
| OpenAI o1 | Reasoning models | https://openai.com/index/introducing-openai-o1-preview | Generative AI |
| AutoGen Studio | Agent orchestration | https://github.com/microsoft/autogen | Agentic AI |
| Strategic CoT | Advanced reasoning | https://arxiv.org/abs/2409.03271 | Generative AI |
| RAG Noise | Retrieval robustness | https://arxiv.org/abs/2408.13533 | Generative AI |
| GameGAN | Simulated worlds | https://github.com/nv-tlabs/GameGAN_code | Generative AI |
| Agentic RAG | Time-series RAG | https://arxiv.org/abs/2408.14484 | Generative AI |
| Paper | Focus | Link | Category |
|---|---|---|---|
| The AI Scientist | Automated discovery | https://paperswithcode.com/paper/the-ai-scientist-towards-fully-automated-open | Generative AI |
| ControlNeXt | Video & image control | https://arxiv.org/pdf/2408.06070v2 | Computer Vision |
| RAG-Checker | RAG diagnostics | https://arxiv.org/abs/2408.08067 | Generative AI |
| Paper | Focus | Link | Category |
|---|---|---|---|
| Microsoft SAMBA | Efficient LLMs | https://arxiv.org/pdf/2406.07522 | Generative AI |
| Quantinuum Quixer | Quantum transformers | https://openai.com/index/extracting-concepts-from-gpt-4 | Quantum AI |
| No Language Left Behind | Multilingual translation | https://github.com/facebookresearch/fairseq/tree/nllb | Generative AI |
| Paper / Model | Focus | Link | Category |
|---|---|---|---|
| Meta Llama-3 | Long-context LLMs | https://huggingface.co/papers/2404.19553 | Generative AI |
| Multi-Token Prediction | Faster inference | https://arxiv.org/abs/2404.19737 | Generative AI |
| Phi-3 | Small efficient LLMs | https://arxiv.org/pdf/2404.14219 | Generative AI |
| FrugalGPT | Cost-aware LLM usage | https://portkey.ai/blog/implementing-frugalgpt-smarter-llm-usage-for-lower-costs | Generative AI |
| GNN-RAG | Graph-based RAG | https://github.com/cmavro/GNN-RAG | Generative AI |
- Track emerging AI system trends
- Identify high-impact papers worth exploring
- Map research to real-world system design decisions
- Stay current with agentic + generative AI evolution
This repository reflects a core belief:
AI value is unlocked not just through models —
but through systems that make those models usable, reliable, and scalable
If this repository is valuable, consider starring it to support continued curation of high-signal AI research and advancement of production-grade AI systems.