© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 1208
    Prev
    Next
    INDEX
    Explanations

    The neuron fires strongly on the very first word of new sections or sentences (i.e. tokens appearing immediately after a segment boundary).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Supervisors
    0.43
     Prosecutor
    0.38
    🕓
    0.38
     ambassadors
    0.37
    iomanip
    0.37
    съ
    0.37
     Commissioners
    0.36
    univ
    0.36
     Supervisory
    0.36
     Enforcement
    0.36
    POSITIVE LOGITS
    ﻴ
    0.47
     والع
    0.45
     يا
    0.43
     сроки
    0.42
     ganze
    0.41
    $,
    0.41
    $&
    0.40
     හා
    0.40
     czy
    0.40
     qualche
    0.40
    Activations Density 0.000%

    No Known Activations