© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 9135
    Prev
    Next
    INDEX
    Explanations

    phrases related to economic struggles or performance metrics

    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
     Dead
    -0.18
     dead
    -0.18
    ead
    -0.18
    abit
    -0.15
     empty
    -0.15
    Dead
    -0.15
    osta
    -0.15
     DEAD
    -0.15
    andoned
    -0.15
    lauf
    -0.15
    POSITIVE LOGITS
     sensitive
    0.38
    Sensitive
    0.35
     sensitivity
    0.33
    -sensitive
    0.33
     vulnerable
    0.32
     Vulner
    0.30
     vulnerability
    0.28
     delic
    0.27
    ensitive
    0.27
     facing
    0.26
    Activations Density 0.059%

    No Known Activations