Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 3555
    Prev
    Next
    INDEX
    Explanations

    references to color and formatting attributes in a graphical context

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

    "color" and diagram code

    np_acts-logits-general · gemini-2.0-flash

    color values assigned

    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
    tenths
    -0.50
     scattata
    -0.49
     typelib
    -0.49
     ruinas
    -0.46
    oa̍t
    -0.45
     direta
    -0.44
     borbo
    -0.43
     dovre
    -0.43
     própri
    -0.43
     termica
    -0.41
    POSITIVE LOGITS
    ValueStyle
    0.57
     Ανακτήθηκε
    0.45
     nonUne
    0.45
    作
    0.44
    tikzpicture
    0.42
     differential
    0.41
     ister
    0.41
     Wagner
    0.40
    oken
    0.40
     حقوق
    0.40
    Activations Density 0.047%

    No Known Activations