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. 90077
    Prev
    Next
    INDEX
    Explanations

    technical terms related to chemical processes and their effects in scientific contexts

    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
    MemoryWarning
    -0.34
    解
    -0.33
     open
    -0.33
    OpenHelper
    -0.33
     Open
    -0.32
     Ос
    -0.31
     liber
    -0.31
     liberating
    -0.31
     freeing
    -0.30
     throw
    -0.30
    POSITIVE LOGITS
     tighter
    0.65
     fermée
    0.63
     tightened
    0.62
    穩定
    0.61
     stabilizing
    0.61
     stabilized
    0.60
     calme
    0.59
    Халык
    0.59
     tight
    0.59
     stability
    0.59
    Activations Density 0.297%

    No Known Activations