# Agent Drift Mitigation on Rath 847 workflows

Source: [Agent Drift Mitigation on Rath 847 workflows](<https://www.sota2.com/research/sota/agent-drift-mitigation-on-rath-847-workflows>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-04-28T10:37:31.966Z

Evaluation of models on the Rath 847 workflows dataset, specifically measuring the effectiveness of Agent Drift Mitigation techniques, quantified by metrics like Drift Reduction.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Agent Drift Mitigation |
| Dataset (including any stated split) | Rath 847 workflows |
| Metric | Drift Reduction |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 4 of 4 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Drift Reduction](<https://www.sota2.com/research/sota/agent-drift-mitigation-on-rath-847-workflows.md?metric=Drift%20Reduction>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Drift Reduction | Result date | Report paper |
| --- | --- | --- | --- |
| Combined | 81.5 | Not recorded | 69f03f5d55a3c32af89fa3dd |
| Adaptive Anchoring | 70.4 | Not recorded | 69f03f5d55a3c32af89fa3dd |
| Drift-Aware Routing | 63 | Not recorded | 69f03f5d55a3c32af89fa3dd |
| Memory Consolidation | 51.9 | Not recorded | 69f03f5d55a3c32af89fa3dd |

### Combined

Drift Reduction: 81.5. Result ID: 69f08dd847f34ce924d73e87.

[Reporting paper](<https://www.sota2.com/research/paper/69f03f5d55a3c32af89fa3dd.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695f39fa53fdd5c36174405e.md>)

Recorded evaluation configuration: {"B(x) Term":"All B(x)","RiskGate Equivalent":"Full pipeline"}. This record does not assert independent reproduction.

### Adaptive Anchoring

Drift Reduction: 70.4. Result ID: 69f08dd847f34ce924d73e86.

[Reporting paper](<https://www.sota2.com/research/paper/69f03f5d55a3c32af89fa3dd.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695f39fa53fdd5c36174405e.md>)

Recorded evaluation configuration: {"B(x) Term":"U(x) threshold","RiskGate Equivalent":"D-UCB1 + Bayesian references"}. This record does not assert independent reproduction.

### Drift-Aware Routing

Drift Reduction: 63. Result ID: 69f08dd847f34ce924d73e85.

[Reporting paper](<https://www.sota2.com/research/paper/69f03f5d55a3c32af89fa3dd.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695f39fa53fdd5c36174405e.md>)

Recorded evaluation configuration: {"B(x) Term":"U(x) stability","RiskGate Equivalent":"Kill Switch + failsafe thresholds"}. This record does not assert independent reproduction.

### Memory Consolidation

Drift Reduction: 51.9. Result ID: 69f08dd847f34ce924d73e84.

[Reporting paper](<https://www.sota2.com/research/paper/69f03f5d55a3c32af89fa3dd.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695f39fa53fdd5c36174405e.md>)

Recorded evaluation configuration: {"B(x) Term":"U(x)","RiskGate Equivalent":"Dirichlet + sliding windows"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.