© 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
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-2B
    7. Residual Stream - 16k
    9. 18-GEMMASCOPE-RES-16K
    11. 8410
    Prev
    Next
    INDEX
    Explanations

    references to crime and violence, particularly involving murderers and their capture

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

    Was/is verb followed by a word indicating destruction or defeat

    np_acts-logits-general · gemini-2.0-flash

    is defeated or destroyed

    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
    出版年
    -0.71
    OGND
    -0.69
     Addi
    -0.65
    utiny
    -0.61
    */].
    -0.59
    Vidite
    -0.58
    ósito
    -0.56
     BIL
    -0.55
    omorphism
    -0.55
     BorderRadius
    -0.55
    POSITIVE LOGITS
     defeated
    0.88
     defeat
    0.71
     vanquished
    0.68
     killed
    0.64
     eliminated
    0.64
     stopped
    0.63
     neutralized
    0.61
     defeats
    0.59
     knocked
    0.59
     caught
    0.59
    Activations Density 0.296%

    No Known Activations