© 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
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 9741
    Prev
    Next
    INDEX
    Explanations

    components related to automotive engineering and diagnostics

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

    technical terms related to systems, technology, and equipment.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    technical, domain-specific nouns denoting systems, instruments, or formal processes in scientific/engineering contexts

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    technical terminology related to mechanical systems, medical procedures, and scientific equipment.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
     bronce
    -0.44
     Erhaltung
    -0.40
     olmayan
    -0.40
     bandeira
    -0.39
    peque
    -0.38
     fréquence
    -0.38
     Verhaltens
    -0.37
     Widers
    -0.36
     explotación
    -0.36
    smaller
    -0.36
    POSITIVE LOGITS
    
    0.59
     مشين
    0.58
     snippetHide
    0.57
    transQ
    0.57
     createSlice
    0.57
    NUMX
    0.54
    новништво
    0.52
    httphttps
    0.52
     Réponses
    0.52
     RER
    0.51
    Activations Density 1.021%

    No Known Activations