© 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. 77715
    Prev
    Next
    INDEX
    Explanations

    titles of books and works

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

    The neuron strongly activates on front‐matter and structural metadata tokens—i.e. titles, headings, section labels, and other all‐caps or numeric entries in document headers and tables of contents.

    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
     like
    -1.00
     cambió
    -0.98
     asemenea
    -0.96
     thereafter
    -0.94
     then
    -0.91
     wtedy
    -0.90
     such
    -0.90
    の名前
    -0.89
     răsp
    -0.87
     should
    -0.86
    POSITIVE LOGITS
    Containing
    1.05
     Containing
    1.02
     mentales
    1.02
     متعلقه
    1.02
     ...,
    0.98
     Etc
    0.96
    衿
    0.94
    Ä
    0.93
     with
    0.92
    及其
    0.92
    Activations Density 0.052%

    No Known Activations