© 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. 22-GEMMASCOPE-RES-16K
    6. 6151
    Prev
    Next
    INDEX
    Explanations

    references to abstract concepts or general ideas

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

    "Things" followed by verbs or related words

    np_acts-logits-general · gemini-2.0-flash

    things happening or described

    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
    •Steering experiments: Gemma 2-2B / 10-GEMMASCOPE-RES-16K

    No Comments

    Negative Logits
     gales
    -0.60
     ufe
    -0.58
    ferrer
    -0.55
    DAC
    -0.53
     uſe
    -0.52
    CODES
    -0.52
     Rosales
    -0.51
     счастья
    -0.50
     Perform
    -0.50
    Hollow
    -0.50
    POSITIVE LOGITS
     happening
    0.99
     happened
    0.93
     Thing
    0.89
     things
    0.88
     THING
    0.88
     THINGS
    0.85
     estimés
    0.84
     happens
    0.84
    Things
    0.84
     CreateTagHelper
    0.83
    Activations Density 0.102%

    No Known Activations