© 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
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 28507
    Prev
    Next
    INDEX
    Explanations

    references to previous information

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

    This neuron identifies referring expressions that point to earlier parts of the text (e.g. “above,” “previous,” “aforementioned”).

    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
     évêque
    -1.02
    itados
    -1.00
    estima
    -1.00
     zijne
    -0.99
    ánicos
    -0.98
    zieher
    -0.98
    لباس
    -0.95
    intéress
    -0.95
     déjeuner
    -0.93
     giacca
    -0.93
    POSITIVE LOGITS
     aforementioned
    1.63
     earlier
    1.27
     above
    1.23
     previously
    1.16
     "
    1.14
     again
    1.09
     mentioned
    1.09
     již
    1.02
     previous
    1.02
     “
    0.98
    Activations Density 0.005%

    No Known Activations