© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 40023
    Prev
    Next
    INDEX
    Explanations

    most + distinctive adjective

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

    This neuron detects the use of the superlative marker “most” (as in “the most …”) in adjective or adverb phrases.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    2
    -2.38
     the
    -2.22
    -2.14
     dasarnya
    -2.13
    for
    -2.11
    を買う
    -2.06
    0
    -2.03
    為
    -2.03
    :
    -1.94
    section
    -1.92
    POSITIVE LOGITS
    쩡
    2.75
    猡
    2.38
     είχε
    2.31
    贗
    2.30
    锲
    2.27
     οποία
    2.23
    er
    2.14
     thing
    2.13
    </h4>
    2.09
    したり
    2.08
    Activations Density 0.029%

    No Known Activations