© 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. 8522
    Prev
    Next
    INDEX
    Explanations

    lie followed by location

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

    This neuron responds to occurrences of the verb “lie” (especially “lies” and “lying”) used to indicate the location of something.

    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
    躥
    -2.67
    From
    -2.30
    where
    -2.20
    絰
    -2.13
    捗
    -2.03
     vorigen
    -1.93
     “(
    -1.93
    érape
    -1.91
    o
    -1.91
    
    -1.86
    POSITIVE LOGITS
    \
    2.30
    -
    2.16
    geſ
    1.75
     liegen
    1.74
    一瞬
    1.73
     descubierto
    1.70
     bosques
    1.68
    的车
    1.65
     Each
    1.64
     небольшой
    1.63
    Activations Density 0.009%

    No Known Activations