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

    Neuronpedia

    Jacobian LensNEW
    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. 42757
    Prev
    Next
    INDEX
    Explanations

    safety guidelines related to working near electrical equipment and hazards

    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
    iglia
    -0.15
    ifo
    -0.14
     Intermediate
    -0.14
    inois
    -0.14
    andom
    -0.14
    inis
    -0.14
    phe
    -0.14
     intermediate
    -0.14
    anni
    -0.14
    igue
    -0.13
    POSITIVE LOGITS
     sensitive
    0.30
    -sensitive
    0.26
     active
    0.25
     fragile
    0.24
     areas
    0.24
     delicate
    0.24
    ensitive
    0.22
     ACTIVE
    0.21
    Sensitive
    0.21
     sensit
    0.21
    Activations Density 0.035%

    No Known Activations