© 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. 131057
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on subjective evaluation words or phrases—i.e. tokens that express judgment of how well something “works,” “looks,” or “plays a role,” etc.

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

    state or condition

    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
    Not in Any Lists

    No Comments

    Negative Logits
     in
    -1.11
    Schöne
    -1.08
     here
    -1.06
    完成后
    -1.06
     last
    -1.05
     thánh
    -1.04
     после
    -0.98
     cuando
    -0.98
    Matériau
    -0.98
     intéressant
    -0.98
    POSITIVE LOGITS
     because
    1.03
     those
    1.02
    €“
    1.01
     minyak
    0.97
    oplasmic
    0.96
    lograph
    0.95
    ]-->
    0.93
     fordi
    0.92
     how
    0.92
     what
    0.91
    Activations Density 0.072%

    No Known Activations