© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 116285
    Prev
    Next
    INDEX
    Explanations

    phrases describing places after where

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

    The neuron consistently activates on the word “where,” marking the start of relative‐clause/adverbial 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_34/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
     and
    -2.58
    them
    -1.59
     or
    -1.50
    挲
    -1.41
     of
    -1.38
    )=>{
    
    -1.38
    🥴
    -1.35
    )("
    -1.32
    +",
    -1.31
    €,
    -1.27
    POSITIVE LOGITS
     there
    2.67
     you
    2.63
     we
    2.44
     they
    2.14
     it
    2.02
    Upon
    1.93
    upon
    1.84
    Where
    1.82
    from
    1.80
    Once
    1.52
    Activations Density 0.054%

    No Known Activations