© 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. 41333
    Prev
    Next
    INDEX
    Explanations

    references to political incidents and legal allegations

    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
     Ledger
    -0.17
    iec
    -0.16
     Doll
    -0.15
    beit
    -0.14
    etch
    -0.14
    宽
    -0.14
     Beste
    -0.14
     reli
    -0.13
    arel
    -0.13
    bek
    -0.13
    POSITIVE LOGITS
     nexus
    0.18
     FEMA
    0.17
     scam
    0.16
     Augusta
    0.16
     conn
    0.15
     irregular
    0.15
    ooke
    0.15
     pon
    0.15
    conn
    0.14
     fake
    0.14
    Activations Density 0.046%

    No Known Activations