Research activity; not a hit rate.
Hypotheses & learning
Follow the reasoning.
What BLUM is investigating, what changed, and which claims still need an outcome.
Unavailable
Snapshot timestamp unavailable
Outcome-driven curriculum
Experimental historical strategy selection, compared against declared baselines.
Checking curriculum snapshot
Reading the stored curriculum; this does not start training.
Current investigation
The latest stored research objective.
Unavailable
No current research rationale is available.
HYPOTHESIS
No testable hypothesis is stored.
A hypothesis and a high confidence score are research inputs. Improved net outcomes against a declared baseline are the evidence of progress.
A prediction may have several evaluated horizons.
Stored analysis, not proof of a corrected model.
Changes still need independent validation.
Latest experiment result
Inspect what the source result actually supports.
Unavailable
No completed result is available.
Reported sample—
Inspect performance evidence →Research limits
Constraints remain visible while the engine operates.
No current paper-forward blocker summary is stored.
No additional budget or throughput reason is reported.
Next experiments
The visible stored queue; selection is separate from execution.
No queued experiment is visible
The next research step has not been stored in this snapshot.
Patterns under investigation
Source-reported findings still require prospective validation.
No positive pattern is stored
The workspace does not substitute generated explanations for measured results.
Rejected patterns
Negative evidence is part of the research record.
No rejected pattern is stored
The workspace does not substitute generated explanations for measured results.
Recent learning runs
Activity timestamps are distinct from effectiveness.
No learning run is visible
A missing timeline does not establish that a worker is active.
Advanced research diagnostics
Inspect the existing experiment and generation evidence in Evolution.