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
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 104929
    Prev
    Next
    INDEX
    Explanations

    instances of programming or data structure elements, especially in the context of input/output operations

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

    XML/programming tags (particularly read/write operations) and punctuation marks in structured data formats.

    eleuther_acts_top20 · claude-3-7-sonnet-20250219Triggered by @lu-christina
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114
    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
     stanovnika
    -0.35
     thème
    -0.34
    üyada
    -0.33
    Życiorys
    -0.33
     symboles
    -0.32
     maîtres
    -0.31
     Mejía
    -0.31
     metodología
    -0.31
    Diwedd
    -0.31
     âmes
    -0.30
    POSITIVE LOGITS
     again
    0.78
    again
    0.71
    Again
    0.70
     Again
    0.68
     AGAIN
    0.66
     novamente
    0.60
     utafitiHapana
    0.59
     다시
    0.54
     опять
    0.54
     weer
    0.53
    Activations Density 0.199%

    No Known Activations