© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 65746
    Prev
    Next
    INDEX
    Explanations

    work by preceding words

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

    Neuron 4 selectively fires on the key content words in that central “progress‐principle” explanation sentence—i.e. terms around “in meaningful work” and contrasting “setbacks in the work.”

    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
     möjlighet
    -0.81
    нення
    -0.76
    btnBack
    -0.75
    wald
    -0.73
    itored
    -0.73
     қ
    -0.71
    processable
    -0.69
    EDI
    -0.69
    InkWell
    -0.69
    得上
    -0.68
    POSITIVE LOGITS
     work
    8.13
    work
    4.34
     trabajo
    4.22
    工作
    4.13
     работы
    4.06
     работе
    4.03
     lavoro
    3.94
    Work
    3.89
     trabalho
    3.83
     работу
    3.81
    Activations Density 0.188%

    No Known Activations