© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 9509
    Prev
    Next
    INDEX
    Explanations

    is or hasis or hasis or hasthis is or hasthis is or hasthis is or has

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

    This neuron strongly activates on the names of technical components or options (e.g. “site,” “spacer,” “option,” “valve,” “switch”) when they’re presented as the key subject of a descriptive sentence.

    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
     of
    -2.23
     this
    -1.48
     these
    -1.37
     milano
    -1.29
     такую
    -1.28
     here
    -1.26
     связанные
    -1.26
     deutschland
    -1.25
    r
    -1.23
    no
    -1.22
    POSITIVE LOGITS
     trasera
    1.50
    ⚥
    1.43
    }\}$
    1.41
    RotationOrder
    1.41
    liana
    1.41
    valves
    1.40
    ()]);
    1.39
    昨天
    1.38
     sekarang
    1.38
    <bos>
    1.38
    Activations Density 0.411%

    No Known Activations