© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 7-GEMMASCOPE-TRANSCODER-16K
    7. 13025
    Prev
    Next
    INDEX
    Explanations

    code snippets or, failing that, racist comments

    oai_token-act-pair · gemini-2.0-flash

    code and conversational text

    np_max-act · gemini-2.0-flash

    say "word-start (▁)" Method used: 2 — because tokens mostly start with ▁ — say "word-start (▁)"

    np_max-act-logits · gpt-5-miniTriggered by @chenshw0109
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_7/width_16k/average_l0_70
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.7.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     idéia
    -0.69
    windowFixed
    -0.69
     enfans
    -0.65
    ykite
    -0.63
     Мексичка
    -0.63
     bevis
    -0.62
    
    -0.62
     répondu
    -0.60
     giustizia
    -0.59
     ainfi
    -0.59
    POSITIVE LOGITS
    ArgsConstructor
    0.69
     (
    0.57
    DispatchToProps
    0.56
    CWE
    0.53
     везе
    0.50
     —
    0.49
     (!_
    0.48
     consultato
    0.48
    PrototypeOf
    0.47
     versa
    0.47
    Activations Density 1.202%

    No Known Activations