© 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. 1710
    Prev
    Next
    INDEX
    Explanations

    phrases related to production processes and their applications

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

    "as [diversity word]" or "wide variety/range"

    np_acts-logits-general · gemini-2.0-flash

    wide variety or range

    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
     HasFactory
    -0.69
    guien
    -0.62
    <bos>
    -0.61
     Référence
    -0.61
    sizeCache
    -0.59
    eseorang
    -0.59
    nsf
    -0.59
    BufferException
    -0.58
    letoe
    -0.57
    OrWhiteSpace
    -0.57
    POSITIVE LOGITS
     diverse
    1.89
     varied
    1.70
    diverse
    1.59
     variety
    1.44
     Diverse
    1.42
    Diverse
    1.40
     different
    1.30
     various
    1.29
    variety
    1.24
    various
    1.20
    Activations Density 0.453%

    No Known Activations