© 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. 41313
    Prev
    Next
    INDEX
    Explanations

    names and titles related to people and events in competitive contexts

    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
    -&
    -0.14
    lobs
    -0.14
    .lex
    -0.12
    472
    -0.12
    VERR
    -0.12
     tabPage
    -0.12
    aç
    -0.12
     bdsm
    -0.12
    ัà¸Ĵ
    -0.12
    лаÑģ
    -0.11
    POSITIVE LOGITS
     amongst
    0.23
    ãģĿãģĹãģ¦
    0.21
     ê·¸ë¦¬ê³ł
    0.21
     among
    0.19
    以åıĬ
    0.19
    among
    0.18
    ï¼Į以åıĬ
    0.18
     à¹ģละ
    0.18
     etc
    0.17
     oraz
    0.17
    Activations Density 0.301%

    No Known Activations