Feedback-driven neural network optimization
Interactive Training: Feedback-Driven Neural Network Optimization, by Wentao Zhang, Yang Young Lu, and Yuntian Deng, makes an ongoing neural-network training process controllable by humans and AI agents. First released on October 2, 2025 and published at EMNLP 2025 System Demonstrations, it exposes optimizer settings, training data, and checkpoints through a control server. Interactive Training 2 extends this approach to a reusable protocol for live training, released on arXiv on July 17, 2026.
Original research
Interactive Training: Feedback-Driven Neural Network Optimization
Wentao Zhang, Yang Young Lu, Yuntian Deng · EMNLP 2025 System Demonstrations
Interactive Training 2: Auditable Control Plane for Live Model Training
Wentao Zhang, Xuanhe Pan, Han Zhou, Yang Lu, Yuntian Deng · arXiv
Work through an example
A browser simulation of requested versus applied controls. It does not launch or modify a training job.
No pending action
Step 10 · learning rate 0.001 · no pending action
A training dashboard shows loss curves and evaluations. Interactive control connects those observations to changes in the process that is producing them. A researcher sees instability, changes the learning rate, and observes subsequent updates in the same run. An LLM controller uses the same feedback cycle: inspect evidence, request an intervention, and inspect its effect.
This differs from choosing a configuration before launch or proposing a new experiment after a run finishes. The active model and optimizer continue while exposed controls change. A checkpoint restore or a new round is a separate operation with its own state transition; the interface should record which operation actually occurred.
Sources: Interactive Training: Feedback-Driven Neural Network OptimizationInteractive Training 2: Auditable Control Plane for Live Model Training
The original framework uses a control server to mediate between a human or automated agent and the training process. Its Trainer integration exposes commands for optimizer hyperparameters, data handling, and checkpoints. A live interface lets the controller observe training and submit changes instead of editing a static recipe and relaunching the program.
The paper presents three case studies spanning training stability, sensitivity to initial hyperparameters, and changing user needs. Its LLM-driven learning-rate example demonstrates an agent reading feedback and intervening during optimization. The system contribution is the connection among live observation, a control interface, and the training loop—not a replacement for the optimizer’s parameter updates.
Sources: Interactive Training: Feedback-Driven Neural Network Optimization
LiveTune predates Interactive Training and exposes LiveVariables for changing optimization parameters without restarting the loop. It supplies an earlier mechanism for live parameter adjustment. Interactive Training’s focus is the human and AI-agent workflow around training, and v2 makes the control vocabulary and event record reusable across applications.
AI Training Manager studies a bounded LLM supervisor that asynchronously reads telemetry and adjusts a running recipe. Its experiments introduce specific failure modes in supervised learning and reinforcement learning. This is a useful comparison at the controller level: what evidence reaches the agent, which actions it can request, and how later outcomes affect its decisions. Infrastructure support and measured controller performance are distinct contributions.
Sources: LiveTune: Dynamic Parameter Tuning for Feedback-Driven OptimizationAI Training Manager: Bounded Closed-Loop Control of Adaptive Training RecipesInteractive Training: Feedback-Driven Neural Network OptimizationInteractive Training 2: Auditable Control Plane for Live Model Training
Expose controls that the training application can actually apply, and define when changes take effect. Record the observation that motivated a request, the requested value, the applied value, the step of application, and the subsequent measurements. A submitted request and a successful update are different events. Keeping them linked makes the behavior inspectable.
Evaluate an automated controller against a stated training objective under explicit workloads and budgets. Count its inference and intervention overhead, preserve the run history, and separate improvements within a continuing run from results of newly initialized rounds. The Interactive Training papers and implementation provide the machinery for this feedback loop; the controller and task determine what the loop accomplishes.
Sources: Interactive Training: Feedback-Driven Neural Network OptimizationInteractive Training 2: Auditable Control Plane for Live Model TrainingAI Training Manager: Bounded Closed-Loop Control of Adaptive Training Recipes
EMNLP 2025 System Demonstrations
An open-source framework in which humans and AI agents intervene in ongoing neural-network training through a control server, changing optimizer settings, training data, and model checkpoints.
The original Interactive Training system connects live monitoring with feedback-driven control of a running training process.
arXiv
A reusable protocol with typed settings, structured actions, application-selected control points, and an ordered journal shared by humans, scripts, and LLM controllers.
Extends the 2025 framework from a Trainer-specific command interface to a common control protocol across training workflows.
arXiv
LiveVariables expose optimization parameters through designated system ports for real-time adjustment without restarting the optimization loop.
An earlier system for changing running optimization processes; compare its parameter mechanism with Interactive Training’s human/agent command interface and v2 protocol.
arXiv
An asynchronous LLM supervisor observes training telemetry and proposes constrained changes to a running training recipe, with experiments on deliberately induced supervised-learning and reinforcement-learning failure modes.
Studies an automated controller over active training, complementing Interactive Training’s reusable infrastructure for human and automated controllers.