© 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. 33365
    Prev
    Next
    INDEX
    Explanations

    concepts related to social justice and community well-being

    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
    252
    -0.14
    ên
    -0.13
    ัà¸ļม
    -0.13
    ead
    -0.13
     certains
    -0.13
    931
    -0.13
    gee
    -0.13
     certain
    -0.12
    616
    -0.12
    ̣
    -0.12
    POSITIVE LOGITS
     everyone
    0.51
     all
    0.45
     both
    0.45
     everybody
    0.43
    everyone
    0.39
     Everyone
    0.38
    Everyone
    0.37
    both
    0.36
     ourselves
    0.32
     Both
    0.31
    Activations Density 0.140%

    No Known Activations