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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 4427
    Prev
    Next
    INDEX
    Explanations

    references to crafting and creative hobbies

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

    demographic information about age groups and life stages.

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

    text segments heavy with acronyms, numbers, and emphatic punctuation (especially parentheticals and sentence-final markers), typical of technical posts or strongly worded comments.

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

    personal narratives and opinions, particularly in informal online posts, forum comments, and reviews.

    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
    …!
    -0.68
     schonmal
    -0.67
     Alongside
    -0.65
     badass
    -0.63
    Alright
    -0.63
    Alongside
    -0.62
    gnancy
    -0.59
    …?
    -0.59
     offseason
    -0.58
    umblr
    -0.57
    POSITIVE LOGITS
     quelquefois
    0.71
     muß
    0.67
     skall
    0.61
     occorre
    0.56
     müßte
    0.56
     wußte
    0.53
     mußte
    0.53
     erforderlich
    0.52
     Moslem
    0.52
    wußt
    0.51
    Activations Density 1.405%

    No Known Activations