© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 72890
    Prev
    Next
    INDEX
    Explanations

    as a role

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron triggers on self-introductory “As a/an [role or profession],” phrases that identify the speaker’s or writer’s role.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     repens
    -1.13
     mercen
    -1.05
     esf
    -1.01
     contente
    -0.98
     oubli
    -0.97
     aulas
    -0.96
    homonymie
    -0.95
     amus
    -0.94
     obst
    -0.91
     étoile
    -0.91
    POSITIVE LOGITS
     you
    2.25
     we
    1.59
     I
    1.55
     your
    1.29
     mình
    1.16
     drewni
    1.03
     want
    1.03
     potrebné
    1.01
     griega
    1.00
     have
    1.00
    Activations Density 0.026%

    No Known Activations