© 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-2-2B
    4. Residual Stream - 16k
    5. 18-GEMMASCOPE-RES-16K
    6. 6554
    Prev
    Next
    INDEX
    Explanations

    punctuation marks and container-related terms

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

    or characters followed by punctuation

    np_acts-logits-general · gemini-2.0-flash

    'Fort', 'Q', 'As', 'Each', 'early', 'stress', 'rejects', 'really', 'there'

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    EndContext
    -0.91
    MemoryWarning
    -0.84
     Савезне
    -0.84
    دانشنامهٔ
    -0.84
     kaarangay
    -0.83
    setVerticalGroup
    -0.79
    ConstraintMaker
    -0.79
    +#+#
    -0.78
    __":
    
    -0.72
    ^(@)
    -0.72
    POSITIVE LOGITS
     The
    0.89
    The
    0.82
     And
    0.75
    And
    0.71
    †
    0.70
     Of
    0.69
     It
    0.69
    ↵↵
    0.68
    It
    0.64
    In
    0.63
    Activations Density 0.080%

    No Known Activations