Research Protocols

Evaluation methods before conclusions

Draft methods for testing model performance, agent behavior, efficiency, and safety. These briefs define what must be measured before AnovaAI Labs publishes a result.

Protocol Library
6 method briefs
AllModel PerformanceAgent BehaviorEfficiencySafety
Featured ProtocolModel PerformanceProtocol draft · Method outline

Task-Specific Distillation Evaluation

A protocol for comparing a compact task-specific model with its reference model. The evaluation records accuracy, abstention behavior, review load, latency, and cost under the same test conditions.

All method briefs

Model PerformanceProtocol draft

Context Position Sensitivity

A controlled comparison of identical instructions placed at different positions in long contexts. Any result must identify the model, context length, run count, scoring rubric, and observed failure modes.

Method outlineReview method
Agent BehaviorProtocol draft

Review Steps in Multi-Stage Agent Tasks

A protocol for testing whether structured review improves safe task completion. It separates raw completion from verified completion and records latency, cost, human corrections, and unresolved risk.

Method outlineReview method
EfficiencyMethod review

Retrieval Quality and Latency as Context Grows

A method for measuring retrieval relevance and response time as indexed context expands. Results must include the storage backend, index configuration, hardware, cache state, and query mix.

Measurement planReview method
SafetyProtocol draft

Prompt Injection Resistance for Tool-Using Agents

A proposed attack suite for agents that read external content or call tools. The protocol distinguishes detection, containment, incorrect tool use, data exposure, and cases requiring human review.

Test outlineReview method
EfficiencyPlanned

Synthetic Training Sample Quality Review

A planned rubric for assessing generated training samples before they enter a fine-tuning set. It evaluates provenance, duplication, task coverage, label confidence, and performance on a held-out real-world set.

Research questionReview method
Review Gates

Research methodology

These are the minimum gates a future result would need to pass before it could support a recommendation.

Clear Hypotheses

Each brief starts with a testable question tied to a real operating decision.

Documented Inputs

Data sources, preparation steps, sampling methods, and exclusions must be recorded.

Repeatable Methods

The model, configuration, scoring rubric, and run conditions must be specific enough to repeat.

Qualified Reporting

Results must include limitations, failure modes, and the conditions where the conclusion does not apply.

FAQ

Questions about the protocol library

No. These are draft and planned evaluation protocols. A method brief should not be cited as evidence of a result. Completed findings will be labeled separately with their data, conditions, and limitations.
You can link to a protocol as a description of a proposed method, but it should not be represented as a completed study or validated outcome.
There is no fixed publication schedule. A result would only be labeled published if its protocol were complete, its evidence reviewable, and the limits of its conclusion stated clearly.
Use the contact page to propose a protocol review or dataset contribution. We will confirm whether a scoped evaluation is feasible.
The stack depends on the protocol. Any completed result must identify the models, frameworks, evaluation harness, configuration, and relevant infrastructure used for that run.
Ready to Start?

Help review a protocol

Bring a research question, dataset, or evaluation concern. We can define the evidence needed before anyone treats an outcome as reliable.

Fixed-price quote before any work starts
You own 100% of the code
30 days of free support
AI agents, custom software & automation
Fixed-price quotes, no hidden fees
30-day post-launch support