© 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. 8972
    Prev
    Next
    INDEX
    Explanations

    phrases indicating negligence or failure to disclose information

    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
     ÑĢавно
    -0.17
    erialize
    -0.17
    untu
    -0.16
    ÄĽla
    -0.15
    .Dom
    -0.15
    SCALL
    -0.14
    eless
    -0.14
    etxt
    -0.14
    ürk
    -0.14
    uti
    -0.14
    POSITIVE LOGITS
     transparent
    0.29
     disclose
    0.27
     release
    0.27
     disclosure
    0.27
     reveal
    0.26
     transparency
    0.26
     disclosed
    0.25
     Transparency
    0.25
     div
    0.25
     disclosing
    0.24
    Activations Density 0.046%

    No Known Activations