© 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. Julius Han · Gated SAE for Llama3-8B-Instruct
    5. Llama3-8B-IT
    7. Residual Stream
    9. 25-RES-JH
    11. 30166
    Prev
    Next
    INDEX
    Explanations

    details about individuals involved in political or criminal activities

    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
    èĪ
    -0.16
    -fontawesome
    -0.15
    agma
    -0.14
    ControlEvents
    -0.14
    çĭIJ
    -0.14
    .heading
    -0.14
    htable
    -0.13
    DEX
    -0.13
     Judicial
    -0.13
    igne
    -0.13
    POSITIVE LOGITS
     Rutgers
    0.40
     NJ
    0.40
    NJ
    0.37
    nj
    0.34
     nj
    0.34
     Jersey
    0.33
     Newark
    0.28
     Bergen
    0.28
     Princeton
    0.26
     Voor
    0.24
    Activations Density 0.053%

    No Known Activations