© 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. 34-GEMMASCOPE-RES-131K
    7. 63333
    Prev
    Next
    INDEX
    Explanations

    sensory and physical experiences

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

    The neuron detects compound adjectives formed as a noun plus a present‐participle (gerund) (e.g. “mind-boggling,” “nose-tickling,” “life-threatening,” etc.).

    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
     être
    -0.99
    Recep
    -0.98
     remover
    -0.94
     monta
    -0.91
     being
    -0.88
    )—
    -0.87
     past
    -0.87
    being
    -0.86
     assass
    -0.85
     captivated
    -0.85
    POSITIVE LOGITS
     originalen
    1.02
     passende
    1.02
     standardi
    0.90
     stabilizing
    0.89
     clen
    0.86
    锈
    0.85
     Hva
    0.84
     wrench
    0.83
     klokken
    0.83
     betreffende
    0.82
    Activations Density 0.045%

    No Known Activations