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-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 1015
    Prev
    Next
    INDEX
    Explanations

    programming syntax elements and delineators

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

    mathematical or code delimiters

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    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
     OFDb
    -0.47
    المناصب
    -0.47
    chkin
    -0.46
    OGND
    -0.45
     ✭✭
    -0.44
    vell
    -0.43
    licherweise
    -0.43
     晶
    -0.42
    BufferException
    -0.42
    >(),
    -0.42
    POSITIVE LOGITS
     disponibilités
    0.52
    EndContext
    0.42
     combinado
    0.40
    arakhand
    0.39
     telefónica
    0.39
    classnames
    0.38
     medesimo
    0.38
    mergeFrom
    0.38
     vectorielle
    0.38
     snippetHide
    0.38
    Activations Density 0.089%

    No Known Activations