Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    "-ction" endings

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     embl
    0.57
     LDAP
    0.55
     kann
    0.53
     Loki
    0.53
     backlinks
    0.53
     Protok
    0.53
     tecn
    0.53
     naam
    0.53
     kompat
    0.53
     electronically
    0.52
    POSITIVE LOGITS
    jection
    0.93
    jections
    0.84
    ception
    0.77
    ruption
    0.74
    vention
    0.73
    -/
    0.73
    versal
    0.73
    duction
    0.72
    ject
    0.71
    ceptions
    0.71
    Act Density 0.505%

    No Known Activations