© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 30445
    Prev
    Next
    INDEX
    Explanations

    phrases indicating potential dangers or challenges

    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
    likely
    -0.17
     likely
    -0.17
     probably
    -0.17
     mümkün
    -0.15
     Likely
    -0.15
     somehow
    -0.15
    æķ¢
    -0.15
    probably
    -0.15
    irs
    -0.14
    ä¸įäºĨ
    -0.14
    POSITIVE LOGITS
     quite
    0.32
     Quite
    0.22
     very
    0.22
    quite
    0.22
     especially
    0.22
     anywhere
    0.21
     downright
    0.21
    Anywhere
    0.20
     surprisingly
    0.20
     pretty
    0.19
    Activations Density 0.076%

    No Known Activations