Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 80405
    Prev
    Next
    INDEX
    Explanations

    references to numerical values and identifiers

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

    reference numeral

    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
     folha
    -0.45
     cultural
    -0.37
     sfondo
    -0.36
     Theorie
    -0.36
     Kultur
    -0.36
     so
    -0.36
     where
    -0.36
     Dadurch
    -0.36
     sodass
    -0.35
    cartão
    -0.35
    POSITIVE LOGITS
    ंदीखरीदारी
    0.73
     CanadaChoose
    0.72
    fjspx
    0.71
     '\\;'
    0.69
     الحره
    0.69
     يتيمه
    0.68
     gyhoeddwyd
    0.68
     nakalista
    0.67
     الرياضيه
    0.65
     esternos
    0.65
    Activations Density 0.001%

    No Known Activations