© 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. 19606
    Prev
    Next
    INDEX
    Explanations

    who followed by a verb

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

    The neuron fires on verbs that denote someone performing an action (especially when they follow a relative pronoun like “who”).

    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
    Our
    -2.02
    They
    -1.98
    It
    -1.88
    to
    -1.87
    also
    -1.83
    一个
    -1.82
    7
    -1.81
    "]();
    -1.80
    }$,
    -1.77
    )$,
    -1.76
    POSITIVE LOGITS
     this
    2.06
     their
    1.88
     innen
    1.65
     those
    1.59
    掟
    1.55
     when
    1.54
    telah
    1.54
     it
    1.46
     the
    1.45
     which
    1.34
    Activations Density 0.015%

    No Known Activations