© 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
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B
    7. Residual Stream - 16k
    9. 20-GEMMASCOPE-RES-16K
    11. 12177
    Prev
    Next
    INDEX
    Explanations

    identifiers and data structures related to programming and data types

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

    technical jargon and code-like tokens—especially all-caps identifiers with underscores, acronyms, numeric/hex constants, and structured syntax typical of source code, specs, or formulas.

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

    technical code and notation, particularly struct definitions, function parameters, mathematical expressions, and hexadecimal values in programming contexts.

    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
     CreateTagHelper
    -0.97
    GEBURTSDATUM
    -0.71
    tagHelperRunner
    -0.71
    تقاوى
    -0.70
    expandindo
    -0.68
    disposing
    -0.66
    UnitTesting
    -0.66
     سكانية
    -0.66
     виправивши
    -0.65
    verifyException
    -0.64
    POSITIVE LOGITS
     متحده
    0.32
    &#
    0.31
    Rohy
    0.31
     compañ
    0.30
    ...
    0.30
    Derbyniad
    0.29
    เจ
    0.29
     vẻ
    0.28
     للا
    0.27
    dė
    0.26
    Activations Density 17.850%

    No Known Activations