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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 17-GEMMASCOPE-RES-65K
    6. 28465
    Prev
    Next
    INDEX
    Explanations

    technical terms and concepts related to superconductivity and quantum mechanics

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 17-gemmascope-res-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_17/width_65k/average_l0_125
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    Data Type
    float32
    Hook Name
    blocks.17.hook_resid_post
    Hook Layer
    17
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     paja
    -0.54
    تقاوى
    -0.52
     professional
    -0.52
    Conne
    -0.51
     Conne
    -0.50
     Reflex
    -0.50
     Foy
    -0.49
     Navig
    -0.49
     Movement
    -0.49
     Rans
    -0.48
    POSITIVE LOGITS
    MENAFN
    0.72
    :✨
    0.72
     quelcon
    0.72
     OFDb
    0.69
     betweenstory
    0.68
     Walkover
    0.68
     soin
    0.66
     vettoriale
    0.66
    InputBorder
    0.65
     proprement
    0.64
    Activations Density 3.090%

    No Known Activations