© 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. 21174
    Prev
    Next
    INDEX
    Explanations

    expressions of frustration or disturbances in social interactions

    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
    rips
    -0.14
     اÙĦدÙħ
    -0.14
     surrounding
    -0.14
     Bath
    -0.13
     risks
    -0.13
    olith
    -0.13
     panicked
    -0.13
    FP
    -0.13
    hips
    -0.13
    áu
    -0.13
    POSITIVE LOGITS
     bother
    0.60
     bothering
    0.54
     bothered
    0.46
     bothers
    0.41
     trouble
    0.40
     annoyance
    0.39
     nuisance
    0.38
     pest
    0.37
     disturbance
    0.37
     disturb
    0.35
    Activations Density 0.035%

    No Known Activations