© 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. 14174
    Prev
    Next
    INDEX
    Explanations

    code-related syntax, particularly involving function declarations and control structures

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

    Code-related punctuation and symbols

    np_acts-logits-general · gemini-2.0-flash

    numbers, code, and legal terms

    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
     {\
    -0.51
     dal
    -0.47
     Michael
    -0.45
     du
    -0.45
     β
    -0.44
     Sim
    -0.44
     á
    -0.44
     firme
    -0.44
     ring
    -0.43
    etal
    -0.43
    POSITIVE LOGITS
    ########.
    1.52
    :+:
    1.45
    libft
    1.04
    #+#
    1.02
     kasarigan
    1.01
    Билгалдахарш
    0.86
     snippetHide
    0.85
     EconPapers
    0.78
     oprot
    0.78
     Efq
    0.78
    Activations Density 0.022%

    No Known Activations