© 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. 22-GEMMASCOPE-RES-131K
    4. 69039
    Prev
    Next
    INDEX
    Explanations

    mounting case, ref_id, first record, short_name, time refreshes

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

    The neuron is keyed to technical, often compound, hardware or device–related terms (e.g. “electro-optical,” “mounting,” “refresh”) commonly found in patent-style descriptions.

    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
     with
    -1.24
     their
    -1.14
    谞
    -1.05
    jadikan
    -1.02
    bibit
    -0.95
    kredit
    -0.93
    bullets
    -0.91
    zany
    -0.90
     Tasche
    -0.90
     used
    -0.88
    POSITIVE LOGITS
    ційний
    1.06
    Дополнитель
    1.04
    Более
    1.00
    son
    1.00
    fficients
    0.98
     (~(
    0.98
    ances
    0.98
    ant
    0.97
    了一個
    0.97
    ieszcz
    0.96
    Activations Density 0.008%

    No Known Activations