© 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. 756
    Prev
    Next
    INDEX
    Explanations

    This neuron fires strongly on single prominent topic or domain words (e.g. “housing,” “fire,” “gambling,” “coffee,” “dementia”) used as section-heading or subject keywords.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    the word "housing" or other specific topic words when they appear multiple times in close succession within a specialized or technical discussion.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    salient topic nouns that denote the main subject or domain of the text (policy areas, fields, or industries).

    oai_token-act-pair · gpt-5Triggered 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
    ,
    0.52
    (?)
    0.51
     two
    0.50
     Da
    0.48
    Da
    0.48
     (
    0.47
    (
    0.46
    、
    0.46
     three
    0.46
    Two
    0.45
    POSITIVE LOGITS
     policymakers
    0.67
     ecosystems
    0.61
     communaut
    0.61
     ética
    0.60
     epistem
    0.60
    ustainability
    0.60
     oamenii
    0.59
     tecnologie
    0.57
     democratization
    0.56
    行业的
    0.55
    Activations Density 0.000%

    No Known Activations