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 - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 42857
    Prev
    Next
    INDEX
    Explanations

    punctuation marks, indicating specific types of list or section breaks in structured text formats

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

    code file paths and imports

    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
     ſtre
    -0.52
     juſ
    -0.51
     ſtand
    -0.49
     deſt
    -0.49
    EndContext
    -0.48
     juſt
    -0.47
    WaitGroup
    -0.47
     accidente
    -0.47
     ſever
    -0.47
     reaſon
    -0.47
    POSITIVE LOGITS
     "..\..\..\
    0.67
     ['./
    0.52
    ("~/
    0.51
    ="./
    0.51
     "..\..\
    0.51
     ویکی‌پدیای
    0.50
    __(/*!
    0.50
     "@/
    0.50
    ('./
    0.50
     '../
    0.49
    Activations Density 0.003%

    No Known Activations