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. 5-GEMMASCOPE-RES-16K
    6. 9271
    Prev
    Next
    INDEX
    Explanations

    themes of degeneration and corruption

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.5.hook_resid_post
    Hook Layer
    5
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     المعيارى
    -0.72
     sī
    -0.58
    OrBuilder
    -0.57
     Numerade
    -0.53
    prins
    -0.53
    EventLoop
    -0.51
    mige
    -0.51
    niss
    -0.51
     suple
    -0.50
    ActionResult
    -0.50
    POSITIVE LOGITS
     decay
    1.16
     degradation
    1.15
     Decay
    1.10
     degra
    1.05
     decaying
    0.99
     decayed
    0.98
     breakdown
    0.98
     deterioration
    0.97
     degrade
    0.96
    breakdown
    0.94
    Activations Density 0.296%

    No Known Activations