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-2-2B
    4. Residual Stream - 65k
    5. 20-GEMMASCOPE-RES-65K
    6. 418
    Prev
    Next
    INDEX
    Explanations

    structured data formats or coding syntax

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

    blocks of continuous whitespace or specific formatting in code

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

    operational amplifier resistor

    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-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    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
     חיצוניים
    -0.86
    ----</
    -0.71
     通販
    -0.70
    SequentialGroup
    -0.70
     تضيفلها
    -0.65
     EconPapers
    -0.65
     createSlice
    -0.64
     ddelweddau
    -0.63
     whiteColor
    -0.60
    LayoutConstraint
    -0.60
    POSITIVE LOGITS
     einerseits
    0.60
    “
    0.55
     hautes
    0.55
    「
    0.55
    "*
    0.55
    "_
    0.55
     '*':
    0.54
    ToAction
    0.52
    FXML
    0.52
     întâ
    0.52
    Activations Density 0.033%

    No Known Activations