© 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. 1753
    Prev
    Next
    INDEX
    Explanations

    locations and proper nouns related to history and culture

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

    ending in common suffixes

    np_acts-logits-general · gemini-2.0-flash

    brands and place names

    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
    -
    -0.44
     per
    -0.40
    ing
    -0.40
     comp
    -0.38
    ADVERTISEMENT
    -0.38
     образом
    -0.38
    craper
    -0.37
    ly
    -0.37
     care
    -0.36
    .
    -0.35
    POSITIVE LOGITS
    ftagPool
    1.10
     Anſ
    1.09
     Jefus
    1.07
     protoimpl
    1.06
    ſelf
    1.04
    Hozzáférés
    1.03
     ་་
    1.03
     Reſ
    1.01
     iſt
    0.98
    تقاوى
    0.98
    Activations Density 0.841%

    No Known Activations