© 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. 22-GEMMASCOPE-RES-131K
    7. 9660
    Prev
    Next
    INDEX
    Explanations

    conditional clauses and consequences

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

    The neuron activates primarily on numeric tokens (digits, years, percentages) and related numeric contexts 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_22/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
    但却
    -1.03
     Anybody
    -1.00
    然后再
    -0.98
     démoc
    -0.96
    Она
    -0.91
    Более
    -0.91
     Toutefois
    -0.89
    Though
    -0.87
     anybody
    -0.86
    Anybody
    -0.86
    POSITIVE LOGITS
     swojej
    1.05
     WITHOUT
    0.97
    [>
    0.96
    🖒
    0.94
     heilig
    0.91
     wśród
    0.91
    Ặ
    0.91
     oa
    0.91
     PERFECT
    0.90
    rrbracket
    0.90
    Activations Density 0.015%

    No Known Activations