© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 65761
    Prev
    Next
    INDEX
    Explanations

    sections, code blocks, and documentation headers

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

    The neuron fires on high-level structural or metadata labels in documents—things like section headers (“Section”), grouping tags (“group”), package/module names (“graphicx”, “System”), 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_10/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
     but
    -2.20
     only
    -1.98
     at
    -1.88
     become
    -1.88
     with
    -1.85
     and
    -1.84
     was
    -1.82
     says
    -1.78
     encourage
    -1.77
    してしまう
    -1.74
    POSITIVE LOGITS
    </h4>
    2.38
    }{
    2.27
    ing
    2.17
    the
    2.16
     koel
    2.08
    }\
    2.06
    }
    
    2.03
    </sup>
    2.03
    ();
    2.00
    mentes
    2.00
    Activations Density 0.000%

    No Known Activations