© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 27534
    Prev
    Next
    INDEX
    Explanations

    discussing changes and associated conditions

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

    The neuron is firing on relatively uncommon, content-bearing words (i.e. high-information, high-distinctiveness terms) rather than common function words.

    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
     Juillet
    -1.22
    manuelle
    -1.05
    ägg
    -1.05
    artig
    -1.02
    őben
    -1.01
     PROCEEDINGS
    -0.98
     Février
    -0.98
     Giugno
    -0.97
     盗
    -0.97
    usaha
    -0.95
    POSITIVE LOGITS
     
    1.38
     “
    1.18
     (
    1.15
    :
    1.04
     –
    1.00
    _
    0.98
    …
    0.96
     «
    0.90
    いろいろ
    0.90
     things
    0.89
    Activations Density 0.087%

    No Known Activations