© 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. 22-GEMMASCOPE-RES-131K
    4. 130036
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on transitive action verbs (often in imperative or present-tense form) that take a pronoun object like “them” or “it.”

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    them up or down

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    骓
    -1.64
     münchen
    -1.45
     familières
    -1.40
     navideño
    -1.38
     estrenar
    -1.37
     démocr
    -1.36
     decorativos
    -1.35
     of
    -1.31
     (
    -1.30
     церков
    -1.29
    POSITIVE LOGITS
     up
    2.02
     out
    1.92
     into
    1.65
    ꔹ
    1.48
     down
    1.45
    carus
    1.38
    igth
    1.37
    N
    1.35
     разных
    1.34
     різних
    1.34
    Activations Density 0.023%

    No Known Activations