© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 4410
    Prev
    Next
    INDEX
    Explanations

    technical terms and abbreviations related to programming and data processing

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

    technical configuration parameters and code macro definitions.

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered by @neilrathi

    uppercase technical acronyms/abbreviations (especially short all-caps signal or parameter names) in code or electronics contexts.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    abbreviated variable names or constants in programming code, particularly those with underscores separating components.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408
    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
     pleaſure
    -0.71
     Majefty
    -0.71
    Tembelea
    -0.51
     étag
    -0.50
     fubject
    -0.50
    wouldn
    -0.49
    SocketChannel
    -0.49
    ziplin
    -0.48
     difp
    -0.47
     deſt
    -0.46
    POSITIVE LOGITS
    s
    1.15
    es
    0.99
    bs
    0.94
    os
    0.90
    ses
    0.89
    ens
    0.88
    zs
    0.86
    ks
    0.86
    as
    0.84
    rs
    0.84
    Activations Density 1.170%

    No Known Activations