© 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
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 24053
    Prev
    Next
    INDEX
    Explanations

    references to wisdom and memory

    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
    ÏĩÏģι
    -0.18
     mastur
    -0.16
    luet
    -0.16
     ΤοÏħ
    -0.15
    icap
    -0.15
     Îijγγ
    -0.15
    «ĺ
    -0.15
     gdk
    -0.15
    λÏī
    -0.15
    elu
    -0.15
    POSITIVE LOGITS
     Lesbian
    0.23
     Zeus
    0.19
    Äĵ
    0.18
     purple
    0.18
     Py
    0.18
     Jason
    0.18
     Metro
    0.18
     demos
    0.17
     Epic
    0.17
     Fortune
    0.16
    Activations Density 0.011%

    No Known Activations