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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 6139
    Prev
    Next
    INDEX
    Explanations

    professionals, places, or things followed by a comma

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

    The neuron fires on common‐noun tokens (things/people/places), i.e. it detects most ordinary nouns in the text.

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

    words related to populated places, particularly "city", "town", and similar geographic administrative terms.

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

    the middle items in coordinated lists (comma-separated series with conjunctions), especially internal nouns within enumerations.

    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
     និង
    0.69
     そして
    0.64
     그리고
    0.63
     তিনটা
    0.60
     陳
    0.59
     thiab
    0.59
    そして
    0.59
     আৰু
    0.59
     選
    0.58
     ஆகியோர்
    0.56
    POSITIVE LOGITS
     cuz
    0.68
    children
    0.64
     नही
    0.63
    doesn
    0.59
     horribly
    0.58
    あなたが
    0.57
     doesn
    0.57
     popping
    0.56
     mangas
    0.56
     alot
    0.56
    Activations Density 0.029%

    No Known Activations