© 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. 5409
    Prev
    Next
    INDEX
    Explanations

    issues related to errors and mistakes in processes or actions

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

    words and phrases related to errors, mistakes, or unintended actions.

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

    descriptions of mistakes or accidental/unintended actions, especially human or clerical errors and incorrect selections.

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

    mistakes or errors, particularly when described or acknowledged (such as accidental actions, incorrect information, or clerical errors).

    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
     enrich
    -0.35
    ">+
    -0.34
     ride
    -0.34
     enriched
    -0.34
    ride
    -0.34
     Knowles
    -0.33
    amas
    -0.33
    unan
    -0.33
    hire
    -0.32
    trina
    -0.32
    POSITIVE LOGITS
     mistakenly
    0.65
    MigrationBuilder
    0.60
     estekak
    0.58
    WebElementEntity
    0.58
     gyhoeddwyd
    0.57
     accidentally
    0.55
     Monfieur
    0.54
     misread
    0.52
     Mistake
    0.52
     inadvertently
    0.51
    Activations Density 0.223%

    No Known Activations