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 - 16k
    5. 9-GEMMASCOPE-RES-16K
    6. 7014
    Prev
    Next
    INDEX
    Explanations

    references to recurrence and repetition

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

    "again" (and variants)

    np_acts-logits-general · gemini-2.0-flash

    again or another

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_9/width_16k/average_l0_88
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.9.hook_resid_post
    Hook Layer
    9
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     conmigo
    -0.45
    alyptus
    -0.44
     onely
    -0.38
     Komunikasi
    -0.37
     cháy
    -0.36
     Presidencia
    -0.36
     Khusus
    -0.36
     sudadera
    -0.35
     kiin
    -0.35
     Sécurité
    -0.35
    POSITIVE LOGITS
    Again
    1.05
    again
    1.04
     Again
    1.04
     again
    1.02
     AGAIN
    0.97
     another
    0.90
    another
    0.87
     erneut
    0.84
    又是一
    0.82
    Another
    0.81
    Activations Density 0.437%

    No Known Activations