© 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-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 100398
    Prev
    Next
    INDEX
    Explanations

    tech brands and entities

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

    proper nouns and brand names, particularly technology companies and product names.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    This neuron fires on capitalized words directly adjacent to numeric data, i.e. tokens appearing immediately before numbers.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ꯸
    -3.08
    因为它
    -2.61
    屃
    -2.59
    antism
    -2.53
     royaltyfri
    -2.44
     believes
    -2.42
    镭
    -2.42
    坼
    -2.38
    而又
    -2.36
    娠
    -2.36
    POSITIVE LOGITS
    ↵↵
    3.16
     
    2.95
     been
    2.84
     contrasting
    2.78
    er
    2.63
     durante
    2.48
     Requires
    2.45
    ه
    2.42
    “
    2.42
    狨
    2.41
    Activations Density 0.043%

    No Known Activations