© 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
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-2B
    7. Residual Stream - 16k
    9. 22-GEMMASCOPE-RES-16K
    11. 5733
    Prev
    Next
    INDEX
    Explanations

    mentions of performance-related actions or tricks

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

    Follows "do/doing" or relates to performance/tricks

    np_acts-logits-general · gemini-2.0-flash

    performing tricks

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.hook_resid_post
    Hook Layer
    22
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    AndEndTag
    -0.62
    FieldBuilder
    -0.61
    PYX
    -0.60
    onedDateTime
    -0.59
    DockStyle
    -0.56
    öße
    -0.55
    omości
    -0.55
     MaterialApp
    -0.53
    RenderAtEndOf
    -0.52
     Thermostat
    -0.52
    POSITIVE LOGITS
     magician
    1.14
     circus
    1.13
     acrobat
    1.08
     Cirque
    1.07
     tricks
    1.07
     Circus
    1.06
     performing
    1.06
     magicians
    1.05
     acrob
    1.05
     performer
    1.03
    Activations Density 0.153%

    No Known Activations