© 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-IT
    7. Residual Stream - 131k
    9. 20-GEMMASCOPE-RES-131K
    11. 98520
    Prev
    Next
    INDEX
    Explanations

    phrases indicating user confirmations or responses

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

    starts of sentences and questions

    np_acts-logits-general · gemini-2.5-flash-lite

    tokens that occur at the start of a turn or sentence—capitalized discourse/response starters (e.g., "This", "There", "Yes", "I", "You", "Thanks") indicating the beginning of an answer or utterance.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    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
     gyhoeddwyd
    -0.53
    iconque
    -0.50
    MigrationBuilder
    -0.48
     typelib
    -0.48
    esgue
    -0.47
    RUnlock
    -0.47
    rawDesc
    -0.46
    يكب
    -0.45
     PagesJaunes
    -0.45
    Spoljašnje
    -0.45
    POSITIVE LOGITS
    bear
    0.44
    halb
    0.41
     IActionResult
    0.40
    izzi
    0.39
    engel
    0.39
    DEA
    0.38
    IMENTAL
    0.37
    MetaObject
    0.37
    дента
    0.36
    roma
    0.36
    Activations Density 0.022%

    No Known Activations