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. 2612
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate support, validation, and positive reinforcement related to personal beliefs and experiences

    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
    PropertyChanging
    -0.65
     <=",
    -0.62
     jScrollPane
    -0.56
    ."));
    -0.54
     تضيفلها
    -0.53
    toHaveBeen
    -0.53
    than
    -0.52
    (");
    -0.52
    AddressBook
    -0.52
     })}
    -0.51
    POSITIVE LOGITS
     regardless
    1.23
     even
    1.10
    regardless
    1.09
     irrespective
    1.07
     without
    0.97
     Regardless
    0.93
    Regardless
    0.91
     despite
    0.88
    even
    0.88
     kahit
    0.84
    Activations Density 0.843%

    No Known Activations