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 - 65k
    5. 20-GEMMASCOPE-RES-65K
    6. 791
    Prev
    Next
    INDEX
    Explanations

    formal elements and structures in mathematical or procedural contexts

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

    the beginning of a new section or entry in the text

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

    fitness and exercise

    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
     وتسجيلات
    -0.57
     doInBackground
    -0.53
    Описание
    -0.51
    istika
    -0.48
     composer
    -0.47
    тут
    -0.47
    ifferentiated
    -0.47
    DISCLAIMER
    -0.46
    Methodology
    -0.46
    LoggerFactory
    -0.45
    POSITIVE LOGITS
     Exercise
    0.97
    Exercise
    0.87
    exercise
    0.86
     exercise
    0.85
     Exercises
    0.83
    Exercises
    0.83
     EXERCISE
    0.78
     exercises
    0.76
     Übung
    0.71
    EXERCISE
    0.70
    Activations Density 0.170%

    No Known Activations