© 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. 72584
    Prev
    Next
    INDEX
    Explanations

    occurrences of formatting cues or structural elements in the text, such as paragraphs and brackets

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

    law citations

    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
    postsleuth
    -0.63
     Bioaccumulative
    -0.61
    UVWXYZ
    -0.60
    createCell
    -0.59
     HttpNotFound
    -0.58
    ⓘ
    -0.56
    噙
    -0.55
    ⟬
    -0.55
     ChromeDriver
    -0.55
    Freund
    -0.54
    POSITIVE LOGITS
     []:
    1.36
     ***!
    0.46
     <<<<<<<<<<<<<<
    0.45
     CreateTagHelper
    0.43
    setVerticalGroup
    0.40
    UnusedPrivate
    0.38
     gastro
    0.38
     rather
    0.37
     large
    0.36
    8
    0.36
    Activations Density 0.006%

    No Known Activations