© 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
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 95395
    Prev
    Next
    INDEX
    Explanations

    specific structures or patterns in the text, particularly focusing on punctuation and contextual markers

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

    listing punctuation and symbols

    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
     surla
    -0.91
    CloseOperation
    -0.76
    parsedMessage
    -0.72
    HostException
    -0.70
     Roskov
    -0.68
    ItemBackground
    -0.68
    BufferException
    -0.65
    ⪜
    -0.63
     doInBackground
    -0.63
    tanleria
    -0.62
    POSITIVE LOGITS
    verifyException
    0.31
     Ведь
    0.29
    <eos>
    0.28
    Teilen
    0.27
     trattato
    0.25
    MENAFN
    0.25
     Kläger
    0.24
    <?
    0.24
    <strong>
    0.24
    larının
    0.23
    Activations Density 0.028%

    No Known Activations