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 - 16k
    5. 21-GEMMASCOPE-RES-16K
    6. 14890
    Prev
    Next
    INDEX
    Explanations

    mentions of actions and states related to belief, knowledge, and transmission of information

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 21-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_21/width_16k/average_l0_70
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.hook_resid_post
    Hook Layer
    21
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     mắn
    -0.59
    دانشنامهٔ
    -0.53
    <bos>
    -0.52
     Pasteur
    -0.52
     تانيه
    -0.50
     Supremo
    -0.44
    webElementXpaths
    -0.43
     laude
    -0.43
    Viitteet
    -0.41
     Lordships
    -0.41
    POSITIVE LOGITS
     nothing
    3.15
    nothing
    2.72
    Nothing
    2.61
     Nothing
    2.55
     NOTHING
    2.49
    NOTHING
    2.21
     nada
    2.17
     rien
    1.99
     nothin
    1.97
     none
    1.87
    Activations Density 0.533%

    No Known Activations