© 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. 8248
    Prev
    Next
    INDEX
    Explanations

    neighbors, adjacent, children

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

    This neuron fires on terms describing adjacency or neighboring relations (e.g., “neighbor,” “neighboring,” “adjacent,” “connected”).

    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
    "]();
    -1.33
    
    -1.30
     about
    -1.28
    alike
    -1.27
     mesmos
    -1.23
    rictive
    -1.23
    sement
    -1.22
    nelse
    -1.20
    ounid
    -1.20
    eners
    -1.18
    POSITIVE LOGITS
    ってない
    1.49
    MICRO
    1.41
     acoso
    1.36
    yyj
    1.24
     behoort
    1.22
     vanuit
    1.17
     other
    1.15
     out
    1.14
     espírit
    1.14
     grupp
    1.13
    Activations Density 0.058%

    No Known Activations