© 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. 2946
    Prev
    Next
    INDEX
    Explanations

    code-related constructs and operations, particularly those involving conditional logic and object manipulation

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

    Abbreviations starting with "ac"

    np_acts-logits-general · gemini-2.0-flash

    ac, acr, acc, or acetyl

    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
    parsedMessage
    -0.64
     Maris
    -0.61
    发表于
    -0.61
    TAWA
    -0.60
     SRP
    -0.57
     Shiv
    -0.57
     Marsh
    -0.57
    RTEE
    -0.54
     Botanic
    -0.53
     Larkin
    -0.53
    POSITIVE LOGITS
     Acc
    2.15
     acc
    2.08
     ACC
    2.04
    Acc
    2.02
    acc
    2.00
     Ac
    1.90
    ACC
    1.86
     AC
    1.81
     ac
    1.74
    Ac
    1.74
    Activations Density 0.380%

    No Known Activations