© 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. 3098
    Prev
    Next
    INDEX
    Explanations

    phrases indicating skepticism or questioning beliefs

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

    Precedes the word "means"

    np_acts-logits-general · gemini-2.0-flash

    explains meaning

    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
    fromnode
    -0.63
    endphp
    -0.55
    PerformLayout
    -0.53
    ftagPool
    -0.52
    qtype
    -0.48
    いかがでしょうか
    -0.43
     ogóle
    -0.43
    Izvori
    -0.41
    ziu
    -0.40
    terne
    -0.40
    POSITIVE LOGITS
     means
    3.58
     mean
    3.31
    means
    3.23
     Means
    3.08
     MEANS
    2.97
    Means
    2.96
     meaning
    2.95
     MEAN
    2.93
     meant
    2.75
     Mean
    2.70
    Activations Density 0.731%

    No Known Activations