© 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. 34-GEMMASCOPE-RES-131K
    7. 76490
    Prev
    Next
    INDEX
    Explanations

    many-to-many and many-to-one relationships

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

    The neuron is tuned to detect occurrences of the substring “many” (e.g. the word “many” itself or as part of compound terms like “ManyToMany”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     of
    -1.09
     rise
    -0.88
    sv
    -0.85
    也
    -0.83
     double
    -0.80
    <i>
    -0.77
    on
    -0.76
     lover
    -0.76
     votre
    -0.74
     instability
    -0.74
    POSITIVE LOGITS
     besø
    1.09
    versary
    1.05
    izieren
    1.03
    禕
    1.02
     chinoise
    1.00
     godk
    0.99
     bekijken
    0.99
     thanks
    0.97
    clerosis
    0.96
     tilby
    0.95
    Activations Density 0.022%

    No Known Activations