© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 32572
    Prev
    Next
    INDEX
    Explanations

    possessive followed by work components

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

    This neuron activates on numeric tokens, detecting mentions of specific numbers (years, counts, stats) in the text.

    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
     їх
    -1.31
     их
    -1.20
     expérience
    -1.02
     país
    -0.99
     their
    -0.99
    sohn
    -0.96
     págs
    -0.95
    歳の
    -0.93
     jejich
    -0.93
    experience
    -0.93
    POSITIVE LOGITS
     contents
    1.57
     inhabitants
    1.38
     creator
    1.37
     occupants
    1.30
     existence
    1.30
     creators
    1.27
     structure
    1.20
     owner
    1.18
     usefulness
    1.14
     predecessor
    1.12
    Activations Density 0.147%

    No Known Activations