© 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. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 3425
    Prev
    Next
    INDEX
    Explanations

    aspects of programming syntax and structure in code

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    code-related tokens like 'int', 'for', and other programming syntax elements.

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered by @neilrathi
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    GEBURTS
    -0.37
     benih
    -0.35
     komunikasi
    -0.33
     colgar
    -0.32
    alone
    -0.32
     actie
    -0.31
     écrivain
    -0.31
    bibitem
    -0.31
     aprobado
    -0.30
     menetap
    -0.30
    POSITIVE LOGITS
    featureID
    0.61
    gameserver
    0.58
    VolleyError
    0.56
    msgTypes
    0.55
    ItemBackground
    0.52
    wantErr
    0.52
    criminator
    0.51
    endpush
    0.51
    prepareStatement
    0.49
     CreateTagHelper
    0.49
    Activations Density 0.066%

    No Known Activations