© 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. 29215
    Prev
    Next
    INDEX
    Explanations

    references to concepts of visibility and elimination within a system

    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
    âĶIJ
    -0.15
    нили
    -0.15
     aloud
    -0.15
    ochen
    -0.15
     Hem
    -0.14
    136
    -0.14
    etchup
    -0.14
     settled
    -0.13
    erli
    -0.13
     flesh
    -0.13
    POSITIVE LOGITS
     bye
    0.25
     Bye
    0.23
    bye
    0.23
     packing
    0.18
     gone
    0.18
     Äiji
    0.17
     permanently
    0.17
     BY
    0.16
    kap
    0.15
     Gone
    0.15
    Activations Density 0.046%

    No Known Activations