© 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. 10-GEMMASCOPE-RES-131K
    7. 100174
    Prev
    Next
    INDEX
    Explanations

    numbers starting with 4

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

    The neuron reliably activates on numeric tokens—especially multi‐digit numbers and decimal values such as volume, page, or citation numbers.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    numeric identifiers and measurements—multi-digit numbers, decimals, ordinals, and citation-style numbers common in legal texts, coordinates, and IDs.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    the digit "4" when it appears in numerical sequences, particularly in legal citations, case numbers, dates, and coordinate data.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     nueve
    -2.47
     других
    -2.22
    ):
    -2.22
     Schauspieler
    -2.22
     recommandée
    -2.17
     otros
    -2.16
     Erinnerungen
    -2.16
    汌
    -2.13
     ส์
    -2.09
     produite
    -2.08
    POSITIVE LOGITS
     
    2.28
     sincero
    2.11
     provocando
    2.09
     anden
    2.06
     baada
    1.98
    ቷ
    1.96
     boul
    1.95
     lakini
    1.95
     hoga
    1.94
    
    1.92
    Activations Density 0.089%

    No Known Activations