© 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. 22-GEMMASCOPE-RES-131K
    7. 35462
    Prev
    Next
    INDEX
    Explanations

    Q:

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

    The neuron selectively activates on numerical tokens and counts (e.g. standalone digits or numbers) 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
     then
    -1.33
     instituciones
    -1.23
    就去
    -1.07
    そこから
    -1.07
     sostenibilidad
    -1.06
    などを
    -1.05
     Unable
    -1.04
     usando
    -1.02
     garantizar
    -1.01
     such
    -1.00
    POSITIVE LOGITS
    Each
    1.38
     Each
    1.15
    Almost
    1.11
    Usually
    1.09
     for
    1.06
    Although
    1.05
    Several
    1.03
    Shortly
    1.02
    Nearly
    1.02
     on
    1.00
    Activations Density 0.001%

    No Known Activations