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 - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 904
    Prev
    Next
    INDEX
    Explanations

    elements related to programming structure and error handling

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

    Precedes URLs, usernames, or legal jargon

    np_acts-logits-general · gemini-2.0-flash

    URL protocols or specific beginnings

    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-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
    ter
    -0.51
     wund
    -0.46
    su
    -0.43
    стта
    -0.43
    la
    -0.41
    ten
    -0.41
    Sub
    -0.41
    s
    -0.40
    ,
    -0.40
     (
    -0.40
    POSITIVE LOGITS
    تقاوى
    0.91
     nakalista
    0.72
     nahilalakip
    0.69
     tartalomajánló
    0.68
     Roskov
    0.68
     ویکی‌پدی
    0.66
     disambiguazione
    0.63
    Personensuche
    0.63
    Portály
    0.61
    তথ্যসূত্র
    0.59
    Activations Density 0.011%

    No Known Activations