© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26
    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 127975
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on promotional comparison language—imperative or descriptive phrases about “finding” or “getting” the cheapest price or best deal (e.g. “find you the cheapest fare by comparing the costs of flights…”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    finding specific information

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     at
    -1.64
     this
    -1.55
     their
    -1.41
     by
    -1.40
     having
    -1.27
     these
    -1.26
     his
    -1.21
     for
    -1.21
     have
    -1.20
     or
    -1.19
    POSITIVE LOGITS
     '',
    
    1.30
     contribuer
    1.28
     aiuta
    1.21
    occupe
    1.20
     reposer
    1.20
    Kenmerken
    1.18
    苙
    1.17
     刺繍
    1.13
     ".";
    1.13
     occuper
    1.13
    Activations Density 0.010%

    No Known Activations