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-2-2B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 2918
    Prev
    Next
    INDEX
    Explanations

    punctuation marks and their relationship to conversational tone and sentiment

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

    Sentences with parenthetical remarks

    np_acts-logits-general · gemini-2.0-flash

    contrast or qualification

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 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
     вместо
    -0.60
     Statt
    -0.57
     zamiast
    -0.54
     invokingState
    -0.53
     Briefly
    -0.52
     Alternatives
    -0.52
    为什么
    -0.52
    Quindi
    -0.52
    <!--[
    -0.51
    Mengapa
    -0.51
    POSITIVE LOGITS
    far
    0.86
     far
    0.85
    Far
    0.76
     Far
    0.74
    Quite
    0.72
     далеко
    0.66
     Quite
    0.66
    KommentareTeilen
    0.64
    FAR
    0.64
     FAR
    0.63
    Activations Density 0.164%

    No Known Activations