© 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. 68487
    Prev
    Next
    INDEX
    Explanations

    consequential punctuation and words

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

    the neuron flags words that describe errors, malfunctions, or other problems in a narrative (e.g. broken, error, stuck, ripped, miscalculation).

    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
    umers
    -0.96
    meisters
    -0.96
    UNIVERSIDAD
    -0.95
    📷
    -0.94
     ceasing
    -0.91
    гласно
    -0.88
    OI
    -0.88
     quoque
    -0.87
    做得
    -0.87
    Terrible
    -0.85
    POSITIVE LOGITS
     so
    1.38
     Fortunately
    1.26
     пришлось
    1.22
    Fortunately
    1.09
     remedy
    1.08
     subsequently
    1.07
     Luckily
    1.05
     sehingga
    1.05
     resulted
    1.04
     then
    1.02
    Activations Density 0.042%

    No Known Activations