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. 18-GEMMASCOPE-RES-16K
    6. 3785
    Prev
    Next
    INDEX
    Explanations

    parentheses and brackets in the text

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

    Code containing parentheses

    np_acts-logits-general · gemini-2.0-flash

    connection, options, difference, union, thread, compile, submit

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    -
    -0.64
     as
    -0.62
     S
    -0.59
     can
    -0.58
     Her
    -0.57
     I
    -0.57
     
    -0.56
     so
    -0.56
     D
    -0.56
     an
    -0.56
    POSITIVE LOGITS
    Vidite
    1.44
    曖昧さ回避
    1.31
     estekak
    1.30
     Roskov
    1.23
     >=",
    1.23
     myſelf
    1.23
     متعلقه
    1.21
    mybatisplus
    1.20
    yntaxException
    1.18
     poffible
    1.17
    Activations Density 0.213%

    No Known Activations