© 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. 18-GEMMASCOPE-RES-16K
    6. 6794
    Prev
    Next
    INDEX
    Explanations

    academic or scholarly references, particularly those related to awards and affiliations in the scientific community

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

    Proceedings of academic societies

    np_acts-logits-general · gemini-2.0-flash

    proceedings of the national academy of sciences

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ソナ
    -0.51
    Ź
    -0.49
    retail
    -0.47
     retail
    -0.46
    ресто
    -0.45
     Retail
    -0.45
    ronesia
    -0.44
     destru
    -0.43
     Lizzie
    -0.43
    Retail
    -0.42
    POSITIVE LOGITS
     academies
    0.88
     Academies
    0.78
     onOptions
    0.78
     Academy
    0.77
    ientí
    0.75
     للمعارف
    0.73
    #+#
    0.73
    ftagPool
    0.73
    Academy
    0.71
     academy
    0.71
    Activations Density 0.236%

    No Known Activations