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. 18-GEMMASCOPE-RES-16K
    6. 15452
    Prev
    Next
    INDEX
    Explanations

    assertive statements related to expectations in testing or verification contexts

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

    Follows "expect" in Javascript tests

    np_acts-logits-general · gemini-2.0-flash

    expect assertions in code

    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
    DebuggerNonUser
    -0.95
    InitVars
    -0.87
    -------------</
    -0.80
    Хьажоргаш
    -0.78
     tartalomajánló
    -0.78
     resourceCulture
    -0.76
    PreExecute
    -0.76
    centralwidget
    -0.74
    ]--;
    -0.73
     nakalista
    -0.72
    POSITIVE LOGITS
    tomie
    0.51
    gnum
    0.50
    Mek
    0.48
    getDrawable
    0.47
    hagic
    0.47
     Gaz
    0.46
     Mess
    0.46
     Ecol
    0.46
     pore
    0.45
    choon
    0.44
    Activations Density 0.021%

    No Known Activations