© 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. 47562
    Prev
    Next
    INDEX
    Explanations

    learn to be strong

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

    The neuron is highly activated by very common, high‐frequency function words (e.g. “a,” “and,” “to,” “with,” “on”).

    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
     or
    -0.91
    ero
    -0.84
    BindingSource
    -0.83
    income
    -0.81
    bio
    -0.79
    引起的
    -0.79
     cuarenta
    -0.79
    inti
    -0.78
    living
    -0.76
     viñ
    -0.76
    POSITIVE LOGITS
     lifelong
    1.26
     confidence
    1.20
    Curios
    1.15
     curiosity
    1.13
     independent
    1.09
     samostat
    1.07
     discover
    1.06
     develop
    1.06
    curios
    1.05
     discovery
    1.04
    Activations Density 0.029%

    No Known Activations