Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 705
    Prev
    Next
    INDEX
    Explanations

    error messages and prompts related to user input issues

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 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
     (
    -0.51
     post
    -0.48
    httphttps
    -0.44
     when
    -0.44
     Bar
    -0.43
     me
    -0.43
     in
    -0.43
     poly
    -0.42
    関連記事
    -0.42
     via
    -0.42
    POSITIVE LOGITS
     please
    1.29
     Please
    1.21
    please
    1.21
    Please
    1.17
     sorry
    1.11
     bitte
    1.03
     pleaſure
    1.02
     plz
    1.02
     PLEASE
    1.02
     Monfieur
    1.00
    Activations Density 0.121%

    No Known Activations