Memory operating system for personalized AI agents. 3-layer architecture (STM, LTM, Knowledge). +37.6% accuracy vs baseline.

Stateful context compression for LLM inference. 128K tokens without degradation. Applicable to long session memory.

Benchmark for long-term memory capabilities of LLM assistants. Opens the path toward more personalized assistants.

+26% accuracy, -91% latency, -90% tokens vs baseline. Persistent structured memory for AI agents.

Review of RAG memory architectures for conversational LLMs. Synthesis of vector DB approaches.

Transition from RAG approaches to long-term memory. Agentic memory management via RL.

Photorealistic talking faces with nuanced expressions. 40 FPS online, 512×512. Not commercialized — risk of incomplete publication.

End-to-end audio-avatar LLM. Emotionally rich facial movements beyond lip-sync. 8B + 0.16B LoRA architecture.

Complete digital human: 3D avatar + expressive speech + grounded dialogue. Rare integrated approach.

Comprehensive review of talking head synthesis techniques. Real-time / expressiveness / quality trilemma documented.

Benchmark for complex style control in TTS. Evaluates 11Labs, Deepgram, OpenAI 4o-mini-TTS.

Personalized and controllable zero-shot spontaneous TTS. Speech style encoder + local prosody encoder.

One-step streaming diffusion for talking avatars. Local-Future Sliding-Window Denoising. Single image + streaming audio → real-time long-form video. Directly applicable to Axis 1 (latency).