© 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. 3993
    Prev
    Next
    INDEX
    Explanations

    large impedance

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

    This neuron fires strongly on distinctive content words—especially less frequent, topic-bearing terms (like “external,” “paid,” or “students”)—i.e. it detects salient domain-specific keywords.

    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
     आणि
    0.75
    力和
    0.72
     instead
    0.69
    さと
    0.67
     и
    0.65
    instead
    0.65
     вместо
    0.65
     invece
    0.63
     или
    0.63
     andre
    0.63
    POSITIVE LOGITS
     alike
    1.26
     উভয়
    0.91
    いずれ
    0.90
     zugleich
    0.83
     উভয়ের
    0.83
     respectivamente
    0.81
     keduanya
    0.75
     উভয়ে
    0.75
     соответственно
    0.75
     それぞれ
    0.75
    Activations Density 0.308%

    No Known Activations