© 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. 22-GEMMASCOPE-RES-131K
    7. 73353
    Prev
    Next
    INDEX
    Explanations

    "Let" followed by an expression

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

    This neuron strongly activates on numeric literal tokens (i.e. 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
    junge
    -1.07
     dekol
    -1.05
     rád
    -1.04
    puede
    -1.03
     sapatos
    -1.02
    vlá
    -1.02
     ако
    -1.01
    fertigen
    -1.00
     možnosti
    -0.99
     gezin
    -0.99
    POSITIVE LOGITS
     now
    1.25
     only
    1.12
     further
    1.09
    ristol
    0.96
    שְׁ
    0.96
     said
    0.94
     by
    0.93
     Does
    0.93
     one
    0.92
     TokenType
    0.92
    Activations Density 0.012%

    No Known Activations