© 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. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 9152
    Prev
    Next
    INDEX
    Explanations

    phrases related to political corruption and investigations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    Juliushanhanhan/llama-3-8b-it-res/blocks.25.hook_resid_post
    Features
    65,536
    Data Type
    float32
    Hook Name
    blocks.25.hook_resid_post
    Hook Layer
    25
    Architecture
    gated
    Context Size
    1,024
    Dataset
    Juliushanhanhan/openwebtext-1b-llama3-tokenized-cxt-1024
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    :
    -0.21
     
    -0.17
     em
    -0.16
     com
    -0.16
    aly
    -0.15
     conveniently
    -0.15
    Âł
    -0.14
    .
    -0.14
     general
    -0.14
    aul
    -0.14
    POSITIVE LOGITS
    raphics
    0.15
     ogs
    0.14
    IDEOS
    0.14
    owell
    0.14
    ãĢĢ↵
    0.14
    odata
    0.13
    uÄį
    0.13
     breat
    0.13
    ä¹ĭä¸Ģ
    0.13
     svaz
    0.13
    Activations Density 0.893%

    No Known Activations