© 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. 22-GEMMASCOPE-RES-131K
    7. 15687
    Prev
    Next
    INDEX
    Explanations

    colors

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

    The neuron is triggered by color names—i.e. it detects mentions of colors in the text.

    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
    pagni
    -0.94
    × 
    -0.91
     verschillen
    -0.91
    ~”
    -0.90
    ớm
    -0.90
     Haziran
    -0.90
     cwt
    -0.89
    diendo
    -0.88
    subreddit
    -0.88
    denas
    -0.87
    POSITIVE LOGITS
    色的
    1.32
    色の
    1.08
     of
    1.08
     instructs
    0.93
     extends
    0.90
     unfolds
    0.90
    复杂的
    0.89
    Trois
    0.88
    來看
    0.87
     finishes
    0.85
    Activations Density 0.126%

    No Known Activations