© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 11477
    Prev
    Next
    INDEX
    Explanations

    references to parasitic relationships and feeding behaviors

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

    Feeds, feeding, or thirst

    np_acts-logits-general · gemini-2.0-flash

    consuming blood or other sustenance

    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
     Gaulle
    -0.46
     发表于
    -0.45
     scher
    -0.41
    Eval
    -0.39
    ByDefault
    -0.39
     guarda
    -0.38
    werker
    -0.38
     sale
    -0.37
    Evalu
    -0.37
     terra
    -0.36
    POSITIVE LOGITS
     feeds
    1.33
     feeding
    1.29
     feed
    1.26
    feeding
    1.18
     Feed
    1.16
     Feeds
    1.12
    feed
    1.12
     insatiable
    1.10
     fed
    1.10
    Feeding
    1.09
    Activations Density 0.209%

    No Known Activations