© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 127917
    Prev
    Next
    INDEX
    Explanations

    The neuron lights up most strongly on very common, high‐frequency tokens (short function words and pronouns).

    oai_token-act-pair · o4-mini
    Triggered by @jyhe0408

    supposed claims about people

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    Negative Logits
    真的好
    -1.05
    ítico
    -0.99
    kernels
    -0.92
    ibição
    -0.92
    യും
    -0.90
    äß
    -0.89
    ⊣
    -0.88
    我先
    -0.86
    我跟
    -0.85
     außergewöhn
    -0.84
    POSITIVE LOGITS
     supposed
    1.50
    supposed
    1.45
     supposedly
    1.41
     supuestamente
    1.35
     pourtant
    1.16
     purported
    1.11
     ought
    0.98
    なのに
    0.97
     should
    0.96
     professed
    0.96
    Activations Density 0.036%

    No Known Activations