© 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. 12414
    Prev
    Next
    INDEX
    Explanations

    phrases related to experiences and events in life

    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
     already
    -0.19
    .timeScale
    -0.18
    already
    -0.17
     Already
    -0.16
     increasingly
    -0.16
     soon
    -0.15
     everywhere
    -0.15
    _already
    -0.15
    porno
    -0.15
    ">//
    -0.15
    POSITIVE LOGITS
     nor
    0.39
     anything
    0.36
     anyone
    0.31
    anything
    0.31
     anybody
    0.27
     Anything
    0.25
     never
    0.25
     Never
    0.25
     any
    0.24
    never
    0.24
    Activations Density 0.078%

    No Known Activations