© 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. 20-GEMMASCOPE-RES-16K
    11. 4060
    Prev
    Next
    INDEX
    Explanations

    technical parameters related to device configuration

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

    configuration parameters and metadata in XML or code-like technical documentation.

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

    numerical values in XML/code parameters, particularly in configuration files.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    numeric literals and assignment-style parameter settings in code/config text, especially small integer flags and attribute/value syntax (including tag closures).

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

    numeric values in structured data or code, particularly in configuration parameters, constants, and assignments.

    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
    出版年
    -0.69
    Personendaten
    -0.67
     propOrder
    -0.67
     للمعارف
    -0.65
     שוליים
    -0.59
    tvguidetime
    -0.54
    +#+
    -0.51
     상세
    -0.49
    ⤹
    -0.48
     كومونز
    -0.47
    POSITIVE LOGITS
     numéros
    0.48
     numéro
    0.43
    ActivityResult
    0.41
    番号
    0.40
     onOptions
    0.39
     números
    0.39
     číslo
    0.38
     number
    0.38
    getNumber
    0.37
     parseInt
    0.36
    Activations Density 0.789%

    No Known Activations