© 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-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 36709
    Prev
    Next
    INDEX
    Explanations

    states, places, or entities

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

    The neuron fires on emotionally intense, visceral or bodily-sensation words (e.g. moan, alone, bones, heart, groans, storm) that convey strong feeling or emotional states.

    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
     To
    -1.10
    what
    -1.08
    How
    -1.01
     On
    -0.97
     между
    -0.96
     There
    -0.96
     BTW
    -0.96
    One
    -0.94
    batas
    -0.94
     barriers
    -0.94
    POSITIVE LOGITS
    了
    1.45
    utrition
    1.23
    thanol
    1.17
    𐐀
    1.16
    來
    1.15
    蹓
    1.15
     manicure
    1.13
    媄
    1.13
     mezzanine
    1.13
     vanguardia
    1.13
    Activations Density 0.028%

    No Known Activations