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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 68753
    Prev
    Next
    INDEX
    Explanations

    various forms of the verb "to be" and its associated descriptive language

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

    is followed by certainty

    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
     pagină
    -0.54
    styleType
    -0.54
     queſta
    -0.51
    adaptiveStyles
    -0.51
     $_"
    -0.50
    脚注の使い方
    -0.50
    habad
    -0.49
     metav
    -0.48
    batore
    -0.47
    middels
    -0.46
    POSITIVE LOGITS
     pretty
    0.35
     priceless
    0.35
     dép
    0.33
    석
    0.33
    SwitchCompat
    0.32
     robar
    0.32
     tæ
    0.32
    ंदीखरीदारी
    0.32
    pretty
    0.31
    AccessorTable
    0.31
    Activations Density 0.023%

    No Known Activations