pub enum CaseKind {
Recall,
KnowledgeUpdate,
Contradiction,
Temporal,
MultiSession,
}Expand description
Classification of an eval case for correctness measurement.
Recall is the default (and the only kind used by existing fixtures).
The new kinds are used by bench/dataset-correctness.json to measure
temporal correctness, contradiction resolution, and knowledge-update quality.
Variants§
Recall
Plain retrieval: the relevant memory should appear in the top-k.
KnowledgeUpdate
A newer memory supersedes an older one; the query asks for current state. The current/correct memory should win; the stale one should NOT appear.
Contradiction
Two memories make contradictory claims; the authoritative one should win.
Temporal
A fact is qualified by an as-of date; the most recent should win.
MultiSession
Multiple sessions produced overlapping memories; the canonical one wins.
Trait Implementations§
Source§impl<'de> Deserialize<'de> for CaseKind
impl<'de> Deserialize<'de> for CaseKind
Source§fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>where
__D: Deserializer<'de>,
fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>where
__D: Deserializer<'de>,
Deserialize this value from the given Serde deserializer. Read more
impl Eq for CaseKind
impl StructuralPartialEq for CaseKind
Auto Trait Implementations§
impl Freeze for CaseKind
impl RefUnwindSafe for CaseKind
impl Send for CaseKind
impl Sync for CaseKind
impl Unpin for CaseKind
impl UnsafeUnpin for CaseKind
impl UnwindSafe for CaseKind
Blanket Implementations§
Source§impl<T> BorrowMut<T> for Twhere
T: ?Sized,
impl<T> BorrowMut<T> for Twhere
T: ?Sized,
Source§fn borrow_mut(&mut self) -> &mut T
fn borrow_mut(&mut self) -> &mut T
Mutably borrows from an owned value. Read more