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-2-2B
    4. Residual Stream - 65k
    5. 16-GEMMASCOPE-RES-65K
    6. 53529
    Prev
    Next
    INDEX
    Explanations

    programming concepts related to arrays and their properties

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

    Danish, specific names, technical terms

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 16-gemmascope-res-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_16/width_65k/average_l0_128
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    Data Type
    float32
    Hook Name
    blocks.16.hook_resid_post
    Hook Layer
    16
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    NOPQRST
    -0.90
     marito
    -0.83
     fratello
    -0.79
    ADR
    -0.78
     bicchiere
    -0.73
    VIAF
    -0.71
     âgées
    -0.70
     McFadden
    -0.68
     Irving
    -0.66
     suivez
    -0.66
    POSITIVE LOGITS
    maline
    0.80
    fontaine
    0.79
     Pali
    0.77
    alyptus
    0.74
    ../../
    0.74
    дент
    0.73
     Danes
    0.73
     Leck
    0.73
     PQ
    0.71
    onomía
    0.71
    Activations Density 0.035%

    No Known Activations