© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 14615
    Prev
    Next
    INDEX
    Explanations

    Ne, outdated, extracted, Daily, img, placed, fortune, tracking

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

    The neuron fires on the final meaningful token in each text segment—i.e. it marks the last content word before a snippet ends.

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

    the word "their" appearing in contexts related to possessions or responsibilities.

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

    markers that express relationships of possession, association, role, or accompaniment between entities

    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
    𝓛
    0.98
    Შ
    0.88
    𝓐
    0.86
     chromospheres
    0.84
    𝓗
    0.84
    맡
    0.83
    ꫝ
    0.83
     代码
    0.82
     торго
    0.81
     orçamento
    0.81
    POSITIVE LOGITS
    í
    0.87
    о
    0.82
    ча
    0.82
     incident
    0.81
    い
    0.80
     inn
    0.78
     epid
    0.77
     at
    0.75
     in
    0.73
    <eos>
    0.73
    Activations Density 0.001%

    No Known Activations