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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 54190
    Prev
    Next
    INDEX
    Explanations

    variations of the word "rew" indicating a focus on renewal or revising actions

    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
    hu
    -0.17
    othermal
    -0.15
     puck
    -0.15
    .Pixel
    -0.15
    hab
    -0.15
    echn
    -0.15
    leted
    -0.14
     Levine
    -0.14
    hang
    -0.14
    lob
    -0.14
    POSITIVE LOGITS
     hat
    0.18
     wing
    0.17
     hy
    0.17
    ound
    0.17
     wings
    0.16
    bsite
    0.16
    istar
    0.16
    irtschaft
    0.15
    anza
    0.15
    issenschaft
    0.15
    Activations Density 0.036%

    No Known Activations