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. 13259
    Prev
    Next
    INDEX
    Explanations

    terms and structures related to mathematical equations and expressions

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

    Mathematical or code-related expressions/symbols

    np_acts-logits-general · gemini-2.0-flash

    planes, affinities, i, Q, Asshug, point

    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
    enterOuterAlt
    -0.47
    addCriterion
    -0.40
    蚪
    -0.38
     gâ
    -0.38
    Datuak
    -0.37
     Traveling
    -0.36
    featureID
    -0.36
    FDRE
    -0.36
     Superhost
    -0.35
    posedge
    -0.35
    POSITIVE LOGITS
    RectangleBorder
    0.55
    AISSEE
    0.46
    BufferException
    0.45
     Tabor
    0.43
     mengal
    0.43
    OrFail
    0.41
     każdej
    0.41
    ंदीखरीदारी
    0.40
    RefNanny
    0.40
    ніципа
    0.40
    Activations Density 1.757%

    No Known Activations