© 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. 34776
    Prev
    Next
    INDEX
    Explanations

    phrases starting with "with"

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

    The neuron strongly activates on the token “with,” marking the introduction of descriptive or subordinate clauses.

    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
     verhind
    -1.36
     油画
    -1.26
     moulin
    -1.24
    
    -1.23
     garcía
    -1.23
     étage
    -1.23
    ﹙
    -1.22
    
    -1.21
     lomba
    -1.20
     westfalen
    -1.19
    POSITIVE LOGITS
     also
    1.54
     It
    1.38
     some
    1.24
    These
    1.23
     {
    1.19
     jednocześnie
    1.19
    Also
    1.17
     Эти
    1.16
     ;
    1.14
     :
    1.13
    Activations Density 0.108%

    No Known Activations