© 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. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B
    7. Residual Stream - 16k
    9. 39-GEMMASCOPE-RES-16K
    11. 94
    Prev
    Next
    INDEX
    Explanations

    recurring themes of constancy and persistence

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 39-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_39/width_16k/average_l0_131
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.39.hook_resid_post
    Hook Layer
    39
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     often
    -0.71
     اصلا
    -0.71
     überhaupt
    -0.66
     artık
    -0.63
     często
    -0.62
     actually
    -0.60
     вовсе
    -0.60
     často
    -0.58
     spesso
    -0.56
     addirittura
    -0.56
    POSITIVE LOGITS
    cky
    0.57
     wondered
    0.56
     been
    0.54
     acompañado
    0.53
     striving
    0.52
    onnaissance
    0.52
    KommentareTeilen
    0.52
     câte
    0.52
     kasarigan
    0.51
     reminding
    0.50
    Activations Density 0.147%

    No Known Activations