© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 131k
    9. 20-GEMMASCOPE-RES-131K
    11. 96633
    Prev
    Next
    INDEX
    Explanations

    symbols and notation commonly used in mathematical contexts

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

    theorems and section references

    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-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
    หมือน
    -0.53
    蜒
    -0.49
     oprot
    -0.48
    يميديا
    -0.47
     CreateTagHelper
    -0.45
    ükemmel
    -0.45
    pulsera
    -0.44
     ویکی‌پدی
    -0.43
     Arquite
    -0.42
    numerusform
    -0.42
    POSITIVE LOGITS
    __":
    0.46
    __':
    0.43
    setVerticalGroup
    0.41
    __*/
    0.40
    ftagPool
    0.38
    
    0.37
    참고
    0.37
     للاسماء
    0.36
    Javadoc
    0.36
    VersionUID
    0.36
    Activations Density 1.448%

    No Known Activations