© 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. 39769
    Prev
    Next
    INDEX
    Explanations

    things, places, or concepts

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

    The neuron fires strongly on salient content nouns or proper names—especially those appearing as key terms at the ends of headings or lines.

    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
     GRATIS
    -1.45
     Obtén
    -1.44
     Meksiko
    -1.43
     rajut
    -1.38
     termurah
    -1.36
    𝟐
    -1.36
     poë
    -1.30
     ſou
    -1.29
     Itali
    -1.28
     élas
    -1.27
    POSITIVE LOGITS
     but
    1.27
     not
    1.25
     like
    1.23
     no
    1.11
     
    1.09
    <?>>
    1.07
     see
    1.05
    <
    1.02
     x
    1.01
     any
    1.00
    Activations Density 0.016%

    No Known Activations