© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 39064
    Prev
    Next
    INDEX
    Explanations

    YouTube videos

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

    The neuron activates on mentions of embedded videos or video links (e.g. “video,” “YouTube,” URLs, and related pointing words).

    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
     your
    -1.22
     process
    -1.20
     create
    -1.16
     will
    -1.13
     data
    -1.09
     our
    -1.08
    TokenNameEQUAL
    -1.02
     make
    -1.00
     czyli
    -0.99
    =
    -0.98
    POSITIVE LOGITS
    YouTube
    1.30
     YouTube
    1.24
    槼
    1.20
    Čes
    1.20
     konserv
    1.18
    kwür
    1.15
    鉺
    1.11
     magnifiques
    1.09
     beijo
    1.09
     officiel
    1.09
    Activations Density 0.031%

    No Known Activations