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. 958
    Prev
    Next
    INDEX
    Explanations

    references to specific dates or the concept of "day."

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

    days and yesterday

    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
    makeConstraints
    -0.57
    oa̍t
    -0.54
     sobran
    -0.54
     esternos
    -0.53
    pecabe
    -0.52
    seamnă
    -0.51
    eloma
    -0.49
     manguera
    -0.48
    retudo
    -0.47
     génie
    -0.47
    POSITIVE LOGITS
    日
    1.03
     日
    0.84
    Portail
    0.73
     Day
    0.67
     day
    0.64
    Day
    0.62
    DAY
    0.57
    httphttps
    0.57
     Sun
    0.56
    일
    0.55
    Activations Density 0.001%

    No Known Activations