© 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 EvalsExportsinterp-engineNEWAPI 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. 22519
    Prev
    Next
    INDEX
    Explanations

    comparisons and assessments of strengths and weaknesses

    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
    ipples
    -0.17
    jem
    -0.14
    dsn
    -0.14
    stakes
    -0.14
     Jeh
    -0.14
    ensem
    -0.13
    -ln
    -0.13
    oplayer
    -0.13
    .gg
    -0.13
    à¤ķन
    -0.13
    POSITIVE LOGITS
     strengths
    0.88
     Strength
    0.83
     strength
    0.82
    strength
    0.78
    Strength
    0.77
    _strength
    0.57
    rength
    0.55
     weakness
    0.50
     weaknesses
    0.46
     stre
    0.44
    Activations Density 0.053%

    No Known Activations