© 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. 34-GEMMASCOPE-RES-131K
    4. 98875
    Prev
    Next
    INDEX
    Explanations

    created by compound words

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

    The neuron fires strongly on hyphenated compound adjectives, particularly those of the form “X-led” (e.g. “pharmacist-led,” “student-led”).

    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
    蔦
    -1.02
    releasing
    -1.00
    reaching
    -0.96
    髙
    -0.93
    ifies
    -0.92
    hinged
    -0.92
    receiving
    -0.92
    zegovina
    -0.91
     bekend
    -0.91
     різних
    -0.90
    POSITIVE LOGITS
     generated
    3.11
     produced
    2.97
     created
    2.52
     constructed
    2.16
     owned
    2.16
     initiated
    2.13
     derived
    2.11
     supplied
    2.11
    produced
    2.08
    generated
    2.03
    Activations Density 0.255%

    No Known Activations