© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 7689
    Prev
    Next
    INDEX
    Explanations

    forgotten words

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

    The neuron fires on content‐bearing words (nouns, main verbs, adjectives, adverbs, pronouns) and ignores light function words.

    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
     it
    -2.36
     of
    -1.94
     then
    -1.81
    2
    -1.59
     if
    -1.52
    4
    -1.48
     bulunan
    -1.41
    6
    -1.41
    7
    -1.36
    1
    -1.35
    POSITIVE LOGITS
     DAMAGES
    1.66
    ----</
    1.58
    vartis
    1.57
    Suomen
    1.50
     kteří
    1.50
    isieren
    1.50
     dano
    1.50
    ização
    1.48
    styleable
    1.48
    gotten
    1.47
    Activations Density 0.006%

    No Known Activations