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-2-9B
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 53806
    Prev
    Next
    INDEX
    Explanations

    occurrences of the article "the" and various punctuation marks in relation to significant phrases or entities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
     rayures
    -0.40
     препратки
    -0.39
     fucking
    -0.38
     larmes
    -0.35
     pretty
    -0.35
     oprot
    -0.35
     algum
    -0.35
    Компания
    -0.35
     ujednoznacz
    -0.35
     shitty
    -0.34
    POSITIVE LOGITS
     communities
    0.56
     المعيارى
    0.54
     ✭✭
    0.53
     ecosystems
    0.51
    SequentialGroup
    0.51
     juſ
    0.50
     philanthrop
    0.50
     nonprofit
    0.49
     pleaſure
    0.48
     subsidi
    0.48
    Activations Density 0.010%

    No Known Activations