© 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. 86899
    Prev
    Next
    INDEX
    Explanations

    establishing relationships or comparisons

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

    The neuron fires most strongly on the main content verb (or key action word) that heads a clause.

    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
    Aufbau
    -1.08
     chevaux
    -1.08
     discul
    -1.07
     bourg
    -1.03
     recens
    -1.03
    Technologie
    -1.03
     textura
    -1.03
     castel
    -1.01
    艄
    -1.00
     cabine
    -0.99
    POSITIVE LOGITS
     of
    1.43
     to
    1.33
     with
    1.31
     towards
    1.23
     on
    1.13
     into
    1.09
     from
    1.02
     or
    1.02
     more
    1.00
     and
    0.97
    Activations Density 0.123%

    No Known Activations