© 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. 68104
    Prev
    Next
    INDEX
    Explanations

    NPR, antitrust, HPLC

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

    This neuron activates for the acronym "NPR" (National Public Radio) when it appears in text.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron is detecting the specific token “NPR.”

    oai_token-act-pair · o4-miniTriggered 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
     commerciales
    -2.27
    耵
    -2.27
    byshire
    -2.19
    '
    -2.17
     actuelles
    -2.16
    tigte
    -2.06
     Dateien
    -2.05
    }:
    -2.05
    Ꮵ
    -2.00
     majesty
    -1.98
    POSITIVE LOGITS
    Ｕ
    2.67
    J
    2.59
    裡的
    2.55
    B
    2.55
    H
    2.52
    W
    2.52
     weitere
    2.45
    V
    2.42
     она
    2.41
     $>
    2.33
    Activations Density 0.005%

    No Known Activations