© 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. 34-GEMMASCOPE-RES-131K
    7. 3277
    Prev
    Next
    INDEX
    Explanations

    specific starting words in sentences

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

    The neuron is tuned to pick out numeric tokens—especially multi-digit numbers and years.

    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
    ContainsKey
    -0.82
     Einsch
    -0.71
     calle
    -0.69
     bailarina
    -0.68
     predic
    -0.67
     pedigree
    -0.67
     Marine
    -0.66
    getElement
    -0.65
     anciens
    -0.65
     européenne
    -0.64
    POSITIVE LOGITS
    婠
    0.91
     martial
    0.90
    Azi
    0.85
    武
    0.74
     APPLICATION
    0.73
     NPCs
    0.73
     Condor
    0.72
    Mur
    0.72
     toiture
    0.72
    Ciri
    0.71
    Activations Density 0.007%

    No Known Activations