© 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. 9384
    Prev
    Next
    INDEX
    Explanations

    saving code and instructions

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

    The neuron tends to fire on infrequent “content” words—rare, non‐stopword tokens rather than common function words.

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

    This neuron fires on salient content nouns (topic-bearing words) in the text.

    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
    적으로
    0.81
     Clique
    0.73
     Cependant
    0.73
    ionar
    0.71
     Shutterstock
    0.71
    otrex
    0.70
    னுடைய
    0.70
     Editar
    0.68
    daki
    0.68
    tip
    0.67
    POSITIVE LOGITS
     escog
    0.85
    ко
    0.84
     envoy
    0.79
     heightened
    0.78
     infliction
    0.78
    з
    0.77
     feces
    0.76
     grossly
    0.76
    ỡng
    0.75
     rebellious
    0.75
    Activations Density 0.000%

    No Known Activations