Results for "shared memory"

Extending agents with long-term memory stores.

Mechanisms for retaining context across turns/sessions: scratchpads, vector memories, structured stores.

Hardware resources used for training/inference; constrained by memory bandwidth, FLOPs, and parallelism.

Training one model on multiple tasks simultaneously to improve generalization through shared structure.

Agents communicate via shared state.

Control shared between human and agent.

Designing AI to cooperate with humans and each other.

Reducing numeric precision of weights/activations to speed inference and reduce memory with acceptable accuracy loss.

An RNN variant using gates to mitigate vanishing gradients and capture longer context.

A system that perceives state, selects actions, and pursues goals—often combining LLM reasoning with tools and memory.

Quantifies shared information between random variables.

Joint vision-language model aligning images and text.

GNN using attention to weight neighbor contributions dynamically.

Decomposing goals into sub-tasks.

Humans assist or override autonomous behavior.

Human controlling robot remotely.

Agents optimize collective outcomes.

Emergence of conventions among agents.

Number of samples per gradient update; impacts compute efficiency, generalization, and stability.

Networks with recurrent connections for sequences; largely supplanted by Transformers for many tasks.

PEFT method injecting trainable low-rank matrices into layers, enabling efficient fine-tuning.

Identifying speakers in audio.

Predicting future values from past observations.

Temporary reasoning space (often hidden).

Dynamic resource allocation.

Running models locally.

Internal representation of environment layout.

System-level design for general intelligence.