© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 12310
    Prev
    Next
    INDEX
    Explanations

    syntax elements and structural markers in programming and data queries

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Negative Logits
    Js
    -0.13
     ëħĦëıĦë³Ħ
    -0.13
    INY
    -0.13
    erosis
    -0.12
    .fm
    -0.12
     UNICODE
    -0.12
    amo
    -0.12
    edd
    -0.12
    .AI
    -0.12
    .Design
    -0.11
    POSITIVE LOGITS
    -LAST
    0.12
     aka
    0.12
     addCriterion
    0.12
    cket
    0.12
     Licensing
    0.12
    acen
    0.12
    readcr
    0.12
     를
    0.12
    APON
    0.11
    TION
    0.11
    Activations Density 0.024%

    No Known Activations