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-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 107939
    Prev
    Next
    INDEX
    Explanations

    phrases concerning randomness and control mechanisms in complex systems

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

    randomness and hidden information

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    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
    PRNewswire
    -0.37
     připo
    -0.37
     poko
    -0.35
     verzorg
    -0.32
     wyda
    -0.32
    bride
    -0.32
     migrantes
    -0.31
     involucrados
    -0.31
    łk
    -0.31
    ]='\
    -0.30
    POSITIVE LOGITS
     random
    0.78
     randomly
    0.72
     RANDOM
    0.70
     Random
    0.68
     randomness
    0.66
    random
    0.64
    Random
    0.63
     secret
    0.60
     hiding
    0.59
     getRandom
    0.59
    Activations Density 0.144%

    No Known Activations