© 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. 10-GEMMASCOPE-RES-131K
    4. 86975
    Prev
    Next
    INDEX
    Explanations

    small talk, great ones, different approaches

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

    The neuron fires on topic- or advice-indicator words (like “tips,” “comments,” “approaches,” “small talk,” “great ones”) and their accompanying prepositions (“on,” “about,” “like,” “of”), i.e. markings of discussion or advice topics.

    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
     had
    -1.77
     never
    -1.56
     didn
    -1.45
     made
    -1.41
    település
    -1.34
     did
    -1.27
     на
    -1.25
    Contactez
    -1.24
    entiated
    -1.24
    人体
    -1.23
    POSITIVE LOGITS
     such
    1.73
     như
    1.41
     SUCH
    1.39
    踯
    1.38
     으로
    1.34
    _+
    1.30
    ⫾
    1.30
    nadequate
    1.30
    such
    1.29
     gmbh
    1.27
    Activations Density 0.092%

    No Known Activations