© 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. 2312
    Prev
    Next
    INDEX
    Explanations

    phrases related to blockages or barriers in various contexts

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

    words related to being stuck, blocked, or experiencing a deadlock.

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered by @neilrathi
    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
     exposure
    -0.38
     exposed
    -0.32
     exposures
    -0.32
     Exposure
    -0.31
     пришли
    -0.30
    ồ
    -0.30
    faker
    -0.29
     wob
    -0.29
    Exposure
    -0.29
     expose
    -0.29
    POSITIVE LOGITS
     stuck
    0.94
     Stuck
    0.91
    Stuck
    0.90
    stuck
    0.89
     stagn
    0.80
     blocked
    0.79
     blockage
    0.78
     stagnant
    0.78
     blocking
    0.77
     stagnation
    0.76
    Activations Density 0.463%

    No Known Activations