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. 1128
    Prev
    Next
    INDEX
    Explanations

    chronological references and historical dates

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

    Years before a specified year

    np_acts-logits-general · gemini-2.0-flash

    ending dates or periods

    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
     grå
    -0.46
     LIABLE
    -0.46
     CURIAM
    -0.46
    bitField
    -0.45
     nahilalakip
    -0.45
     esternos
    -0.44
     Ephesus
    -0.43
    hilangan
    -0.43
     مشين
    -0.42
    ormick
    -0.42
    POSITIVE LOGITS
     until
    0.65
     Until
    0.57
    until
    0.57
    Until
    0.53
     UNTIL
    0.48
     till
    0.47
     עד
    0.46
     hasta
    0.43
     untill
    0.42
    까지
    0.41
    Activations Density 0.547%

    No Known Activations