© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 38997
    Prev
    Next
    INDEX
    Explanations

    increased hippocampal tau phosphorylation

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

    JavaScript return statements within functions.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron activates almost exclusively on numeric literals (e.g. decimal numbers) in code.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     their
    -1.32
     even
    -1.20
     mendengar
    -1.20
     provide
    -1.13
     gorro
    -1.12
    他们
    -1.10
     they
    -1.09
     not
    -1.09
     which
    -1.09
     tirantes
    -1.09
    POSITIVE LOGITS
    ariest
    1.13
    udges
    1.09
     conocido
    1.07
    Kilde
    1.05
    越是
    1.03
    普通的
    1.02
    可能有
    1.02
     uninten
    1.02
     teammate
    1.01
     ayudará
    1.01
    Activations Density 0.000%

    No Known Activations