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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 3929
    Prev
    Next
    INDEX
    Explanations

    code elements related to versioning and defining version properties

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

    programming syntax and code-related tokens, particularly those related to configuration or build processes.

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered by @neilrathi

    code snippets showing added-lines in diffs and formatted string literals (printf/sprintf-style with escapes, format specifiers, and angle-bracketed tags)

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

    formatting patterns in code that involve string literals with escape sequences, particularly after assignment operators or in printf statements.

    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
     termica
    -0.35
     rentrée
    -0.35
     Occidente
    -0.32
     Administración
    -0.32
     nubes
    -0.31
     dueño
    -0.30
    ksanakan
    -0.30
     inauguración
    -0.30
     Jerome
    -0.29
     règle
    -0.29
    POSITIVE LOGITS
    
    0.71
     resourceCulture
    0.66
     autorytatywna
    0.64
    WriteLiteral
    0.61
     الاطلاع
    0.60
     OMITBAD
    0.59
    TargetException
    0.57
    Personensuche
    0.57
    principalTable
    0.56
     виправивши
    0.55
    Activations Density 0.246%

    No Known Activations