© 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. 22-GEMMASCOPE-RES-16K
    6. 7450
    Prev
    Next
    INDEX
    Explanations

    terms associated with vehicle safety ratings and crash testing

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

    airbags, safety, and collisions

    np_acts-logits-general · gemini-2.0-flash

    occupant restraint safety

    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
    rouille
    -0.51
    providedIn
    -0.51
     resourceCulture
    -0.50
    odly
    -0.48
    expandindo
    -0.48
     Winaray
    -0.48
     esplor
    -0.46
    AddTagHelper
    -0.46
     sourdough
    -0.45
     dredging
    -0.45
    POSITIVE LOGITS
     safety
    1.25
     Safety
    1.14
     crash
    1.13
     Crash
    1.10
    Safety
    1.08
    safety
    1.04
     airbags
    1.01
    Crash
    1.00
     SAFETY
    0.95
     airbag
    0.93
    Activations Density 0.179%

    No Known Activations