© 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. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 131k
    9. 20-GEMMASCOPE-RES-131K
    11. 11059
    Prev
    Next
    INDEX
    Explanations

    various types of brackets or quotation marks used in coding and configuration contexts

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

    environment variables and config

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
     juſ
    -0.77
     pouvoit
    -0.73
    majánló
    -0.73
     Inſ
    -0.71
    tagHelperRunner
    -0.69
     auroit
    -0.68
     feroit
    -0.66
     ſon
    -0.66
    iſen
    -0.66
     ब्रेकडाउन
    -0.65
    POSITIVE LOGITS
     "
    0.52
    ifdef
    0.52
    Getenv
    0.50
    ]['
    0.50
    ]?.
    0.48
    ENV
    0.48
     '
    0.48
    ifndef
    0.47
     $
    0.47
    ]->
    0.45
    Activations Density 0.001%

    No Known Activations