© 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. 39279
    Prev
    Next
    INDEX
    Explanations

    list before following

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

    This neuron is highly active on the very first word of each segment or sentence, effectively marking the beginning‐of‐sequence token.

    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
    licante
    -1.10
    vertising
    -1.05
    😱
    -1.02
    鏵
    -1.00
    entification
    -0.99
    ॠ
    -0.98
    😂😂
    -0.98
    銓
    -0.97
    ൌ
    -0.96
     cadáver
    -0.96
    POSITIVE LOGITS
     —
    1.18
     overarching
    1.05
     =
    0.99
    ulho
    0.93
    NO
    0.93
    などを
    0.91
     alluring
    0.90
    NI
    0.89
     famed
    0.88
     to
    0.88
    Activations Density 0.012%

    No Known Activations