© 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. 21416
    Prev
    Next
    INDEX
    Explanations

    concerns and anxieties related to identity and societal expectations

    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
    patial
    -0.15
    Shock
    -0.14
    ArgumentException
    -0.14
    -NLS
    -0.13
    ercul
    -0.13
     Shock
    -0.13
    丸
    -0.13
    urator
    -0.13
    auer
    -0.13
     takdir
    -0.13
    POSITIVE LOGITS
     fear
    0.64
     Fear
    0.58
     fears
    0.58
    Fear
    0.55
     worry
    0.53
     anxiety
    0.49
     worried
    0.48
     worries
    0.46
     fearful
    0.46
     feared
    0.45
    Activations Density 0.055%

    No Known Activations