© 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. 104509
    Prev
    Next
    INDEX
    Explanations

    paths in graphs

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

    references to paths in graphs or networks, particularly in technical or mathematical contexts.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron fires on occurrences of “path” (and its plural “paths”) or related words like “reach” and “ways,” i.e. when the text discusses counting or reaching paths in a graph.

    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
     terkena
    -1.46
     sosok
    -1.41
    ︕
    -1.33
    rupulous
    -1.31
    鹮
    -1.28
    一样
    -1.26
    itorio
    -1.25
    Educación
    -1.23
    льше
    -1.23
     tersenyum
    -1.22
    POSITIVE LOGITS
     that
    1.30
    (
    1.28
    ){$
    1.27
    ſelves
    1.26
     to
    1.24
    :
    1.23
    前面的
    1.22
     path
    1.22
     consentimento
    1.22
     Beyond
    1.21
    Activations Density 0.047%

    No Known Activations