© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 39880
    Prev
    Next
    INDEX
    Explanations

    discourse conjunctions

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

    The neuron lights up on sentence‐level adverbs and discourse markers (e.g. just, unfortunately, moreover, in fact) that signal transitions or speaker attitude.

    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
     with
    -1.76
    щены
    -1.35
    </h5>
    -1.30
     dessutom
    -1.28
     også
    -1.27
     extremadamente
    -1.27
    communist
    -1.25
     causing
    -1.24
     Bearbeit
    -1.23
     myös
    -1.22
    POSITIVE LOGITS
    </strong>
    1.59
    </em>
    1.40
    isation
    1.28
    uegos
    1.27
    
    1.27
    these
    1.23
     moderne
    1.23
    </sup>
    1.23
    紧接着
    1.23
     autres
    1.20
    Activations Density 0.062%

    No Known Activations