Pine AI
Popular repositories Loading
-
programmable-kv
programmable-kv PublicModels Take Notes at Prefill: KV Cache Can Be Editable and Composable (arXiv:2606.17107) — paper, code, results, and interactive companion.
-
user-as-code
user-as-code PublicExecutable memory for personalized agents: represent a user's memory as typed Python state + constraints an interpreter can run, instead of retrieved text. Code, experiments, and reference implemen…
Python 9
-
user-as-engram
user-as-engram PublicA user's facts become a few rows in a content-addressed memory table — not a rewrite of the model. Paper, code, data, and interactive site for "User as Engram" (arXiv:2606.19172).
Repositories
- principal-loyalty Public
Whose Side Is Your Agent On? PrincipalBench — a multi-turn benchmark and post-training methods (prompt scaffold + per-token-KL distillation) for multi-party principal loyalty in LLM agents.
- rlvp Public
Penalize the Path, Reward the Outcome — verifiable per-action penalties as a dense channel for deployable, sample-efficient agentic RL (GRPO). Paper: arXiv:2607.07435
- user-as-code Public
Executable memory for personalized agents: represent a user's memory as typed Python state + constraints an interpreter can run, instead of retrieved text. Code, experiments, and reference implementation for the paper (arXiv:2606.16707).
- user-as-engram Public
A user's facts become a few rows in a content-addressed memory table — not a rewrite of the model. Paper, code, data, and interactive site for "User as Engram" (arXiv:2606.19172).
- aoi Public
AOI: a model-agnostic perception layer giving computer-use agents adaptive video + audio perception with zero retraining, plus DynaCU-Bench (100 dynamic browser tasks). Paper + benchmark + code.
- TalkAct Public
Talking While Acting: a fast/slow dual-model agent that holds a real-time phone conversation while autonomously operating a computer. System, VoiceComputerBench benchmark, and paper.
- namerank Public
NameRank: measuring LLM-mediated recognition as the post-bibliometric impact channel
- interaction-scaling Public
Interaction Scaling: a third inference-time compute axis grounded in environment feedback — proposer–reviewer harness, grounded evaluation, and the paper source.
- transparent-offload Public
Fine-grained computation offload for off-the-shelf servers in tens of lines — paper (arXiv:2607.02630), code, and every measurement. Overlap accelerator offloads (GPU/HSM/inference) with other requests by rerouting through the server's own suspend/resume machinery, plus a zero-edit LD_PRELOAD fiber runtime.
People
This organization has no public members. You must be a member to see who’s a part of this organization.
Top languages
Loading…
Most used topics
Loading…