Results for "changing target relationship"
Techniques that fine-tune small additional components rather than all weights to reduce compute and storage.
PEFT method injecting trainable low-rank matrices into layers, enabling efficient fine-tuning.
Reducing numeric precision of weights/activations to speed inference and reduce memory with acceptable accuracy loss.
Methods to protect model/data during inference (e.g., trusted execution environments) from operators/attackers.
System design where humans validate or guide model outputs, especially for high-stakes decisions.
Forcing predictable formats for downstream systems; reduces parsing errors and supports validation/guardrails.
Methods for breaking goals into steps; can be classical (A*, STRIPS) or LLM-driven with tool calls.
AI focused on interpreting images/video: classification, detection, segmentation, tracking, and 3D understanding.
A theoretical framework analyzing what classes of functions can be learned, how efficiently, and with what guarantees.
A narrow minimum often associated with poorer generalization.
A wide basin often correlated with better generalization.
Capabilities that appear only beyond certain model sizes.
Set of all actions available to the agent.
Models that define an energy landscape rather than explicit probabilities.
Expected cumulative reward from a state or state-action pair.
Maps audio signals to linguistic units.
Expected return of taking action in a state.
Increasing model capacity via compute.
Attention between different modalities.
Increasing performance via more data.
Cost of model training.
Measures joint variability between variables.
Alternative formulation providing bounds.
Model trained on its own outputs degrades quality.
Model relies on irrelevant signals.
Dynamic resource allocation.
Running models locally.
AI systems that perceive and act in the physical world through sensors and actuators.
Field combining mechanics, control, perception, and AI to build autonomous machines.
Hardware components that execute physical actions.