© 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 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. 42602
    Prev
    Next
    INDEX
    Explanations

    references to the challenges of attention and memory in relation to modern technology

    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
    oya
    -0.17
    /*!↵
    -0.16
    loat
    -0.15
     palp
    -0.14
    .ali
    -0.13
    ÏİÏģα
    -0.13
    anh
    -0.13
     Pepper
    -0.13
    arness
    -0.13
    enthal
    -0.13
    POSITIVE LOGITS
     brain
    0.55
     mind
    0.48
     brains
    0.45
    brain
    0.45
     Brain
    0.45
    Brain
    0.44
     minds
    0.42
    mind
    0.36
     Mind
    0.35
    Mind
    0.31
    Activations Density 0.123%

    No Known Activations