Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 61850
    Prev
    Next
    INDEX
    Explanations

    concepts related to advantages and positive changes in various contexts

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

    booms and busts or advantages

    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
    ILAR
    -0.38
    protoimpl
    -0.38
    niająca
    -0.38
     dou
    -0.35
    ervis
    -0.35
     Jar
    -0.35
     STO
    -0.34
     ***
    -0.34
     bud
    -0.34
    zeß
    -0.34
    POSITIVE LOGITS
    PreInfinity
    0.70
    GOTREF
    0.65
    XmlAccessType
    0.63
    posedge
    0.59
    GRANTED
    0.57
    PropertyChanging
    0.53
    Diwedd
    0.52
    mergeFrom
    0.50
    RectangleBorder
    0.50
     or
    0.49
    Activations Density 0.147%

    No Known Activations