© 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-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 13609
    Prev
    Next
    INDEX
    Explanations

    references to personal websites and blogs

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

    before colons, semicolons, or quotes

    np_acts-logits-general · gemini-2.0-flash

    Reader, writer, apologize

    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-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    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
    تقاوى
    -0.56
     nakalista
    -0.56
     незавершена
    -0.52
     محفوظة
    -0.52
    ValueStyle
    -0.51
     vPvB
    -0.50
    الإنجليزية
    -0.48
    🟤
    -0.48
    UnitTesting
    -0.46
    цездатний
    -0.46
    POSITIVE LOGITS
     apologize
    0.35
    Reader
    0.35
     مُعرِّف
    0.35
     Reader
    0.33
    something
    0.33
     apologizing
    0.33
     veggie
    0.32
     apologized
    0.31
     ...
    0.31
     goddamn
    0.31
    Activations Density 0.222%

    No Known Activations