© 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. 58396
    Prev
    Next
    INDEX
    Explanations

    left behind or left in locations

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

    The neuron activates on mentions of items being left or dropped off somewhere (e.g. “left,” “dropped off”).

    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
     maillot
    -1.48
    complish
    -1.45
     Après
    -1.38
    Ｖ
    -1.34
     noires
    -1.30
     luxem
    -1.30
    commod
    -1.28
     pratiquer
    -1.27
     braços
    -1.27
     pourra
    -1.27
    POSITIVE LOGITS
     in
    1.88
     for
    1.76
     on
    1.71
     a
    1.33
     at
    1.30
    Spawner
    1.29
     серия
    1.29
     by
    1.25
     during
    1.20
     ilustración
    1.18
    Activations Density 0.031%

    No Known Activations