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. 240
    Prev
    Next
    INDEX
    Explanations

    text related to licensing and permissions in software documentation

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

    content related to software licensing and permissions

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

    code pragmas and methods

    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
    déric
    -0.50
     Lauter
    -0.50
     NSCoder
    -0.49
     Cæsar
    -0.48
     peony
    -0.48
    ousand
    -0.47
    =$?
    -0.45
     Dox
    -0.44
     penetrated
    -0.44
    urlpatterns
    -0.44
    POSITIVE LOGITS
    SequentialGroup
    0.72
     nakalista
    0.70
    makeConstraints
    0.70
    ConstraintMaker
    0.61
    RenderAtEndOf
    0.60
    deelte
    0.59
    rzost
    0.58
     Normdatei
    0.58
    +#+#
    0.55
    uxley
    0.54
    Activations Density 0.094%

    No Known Activations