© 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. 13487
    Prev
    Next
    INDEX
    Explanations

    Me, R, Ph, CD, black

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

    The neuron responds to very high‐frequency tokens (common function words like “the,” “with,” “containing” and punctuation marks such as dashes).

    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
     排骨
    -1.20
     everybody
    -1.18
     everyone
    -1.13
     ciclista
    -1.12
    nós
    -1.11
    everybody
    -1.10
    everyone
    -1.09
    卺
    -1.09
     our
    -1.07
    Terug
    -1.05
    POSITIVE LOGITS
     only
    1.11
    Me
    1.04
     has
    1.04
    Z
    1.03
    P
    1.00
     apenas
    0.99
    d
    0.98
    En
    0.98
    V
    0.98
    ...
    0.96
    Activations Density 0.000%

    No Known Activations