© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 4868
    Prev
    Next
    INDEX
    Explanations

    works well

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

    The neuron activates specifically on the adverb “well” when it’s used to describe how something works, fits, or contrasts.

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

    the word "well" when it appears in phrases describing how things function, fit, or work together.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    phrases expressing positive evaluations of performance, fit/compatibility, or attractive appearance.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ≣
    0.79
    ای
    0.79
    ي
    0.78
    ди
    0.78
    МА
    0.77
    LEVEL
    0.77
    انا
    0.77
    רא
    0.75
    y
    0.75
    رس
    0.74
    POSITIVE LOGITS
    currentColor
    0.83
    ̣ng
    0.76
    ্যায়
    0.75
     graag
    0.74
    groomed
    0.74
     Absatz
    0.70
    esteem
    0.69
    onds
    0.68
    unteer
    0.68
     gosto
    0.68
    Activations Density 0.020%

    No Known Activations