© 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. 22228
    Prev
    Next
    INDEX
    Explanations

    arguments regarding moral accountability and economic fairness

    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
     поба
    -0.19
    (always
    -0.16
    Å¡ov
    -0.15
    ISM
    -0.15
    uzey
    -0.15
    okit
    -0.15
    Ø´ÙħاÙĦÛĮ
    -0.15
    çı
    -0.14
    Charsets
    -0.14
    iferay
    -0.14
    POSITIVE LOGITS
     why
    0.57
    why
    0.44
     then
    0.43
     Why
    0.41
     shouldn
    0.40
    Why
    0.39
    then
    0.36
     WHY
    0.34
     pourquoi
    0.31
    为ä»Ģä¹Ī
    0.30
    Activations Density 0.046%

    No Known Activations