© 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-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 70864
    Prev
    Next
    INDEX
    Explanations

    office and department names

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

    This neuron detects possessive ’s endings in the names of government or law‐enforcement offices (e.g. “Attorney General’s,” “Sheriff’s Department,” “District Attorney’s Office”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    гора
    -1.35
     også
    -1.30
    自己的
    -1.29
    不好的
    -1.28
     quiere
    -1.28
     Their
    -1.28
     их
    -1.27
    うわ
    -1.26
     myös
    -1.26
     для
    -1.25
    POSITIVE LOGITS
    <bos>
    1.42
     γυναι
    1.37
     frattempo
    1.33
     parco
    1.23
     office
    1.20
     recupero
    1.18
    drums
    1.16
     team
    1.16
     department
    1.14
     tentativo
    1.13
    Activations Density 0.011%

    No Known Activations