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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 15-GEMMASCOPE-RES-16K
    6. 2332
    Prev
    Next
    INDEX
    Explanations

    statistical analysis terminology and related data descriptors

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.15.hook_resid_post
    Hook Layer
    15
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    cioni
    -0.45
     favorites
    -0.40
     mà
    -0.39
    azi
    -0.39
    nero
    -0.37
    Closer
    -0.36
    LEncoder
    -0.36
    eling
    -0.36
    yal
    -0.36
    шту
    -0.35
    POSITIVE LOGITS
     accompanying
    1.11
     afterwards
    1.06
     associée
    0.94
     correspondiente
    0.92
     concomitant
    0.92
     associated
    0.92
     afterward
    0.90
     thereupon
    0.88
     Afterwards
    0.87
     correspondingly
    0.87
    Activations Density 0.707%

    No Known Activations