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. 66990
    Prev
    Next
    INDEX
    Explanations

    actions involving physical manipulation or removal of clothing or body parts, particularly in aggressive or forceful contexts.

    oai_token-act-pair · claude-3-5-sonnet-20240620Triggered by @dribnet

    actions involving tearing or breaking items into smaller pieces or removing clothing

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

    making undone or open

    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
     Lightboxes
    -0.40
     sanitarios
    -0.40
     invité
    -0.39
    Hentet
    -0.36
     Füße
    -0.36
    isateur
    -0.35
     kasarigan
    -0.35
     Савезне
    -0.34
     Infór
    -0.34
     Grüßen
    -0.34
    POSITIVE LOGITS
     tearing
    0.56
     peeled
    0.55
    ScopeManager
    0.54
     peeling
    0.52
     ripping
    0.51
     Inner
    0.51
     раск
    0.51
     ripped
    0.49
     stripped
    0.47
    fastjson
    0.47
    Activations Density 0.131%

    No Known Activations