© 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. 30030
    Prev
    Next
    INDEX
    Explanations

    negative evaluations and critiques related to performance or qualities

    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
     But
    -0.19
    ï¼Į以åıĬ
    -0.17
    ãĢĤãģĿãģĹãģ¦
    -0.16
    But
    -0.16
    ãĢĤä½Ĩ
    -0.15
     äºİ
    -0.15
     ÙĪÙĦÙĥÙĨ
    -0.14
    irl
    -0.14
     etc
    -0.14
    onas
    -0.14
    POSITIVE LOGITS
    :
    0.28
    ;
    0.19
     إذ
    0.18
    ãģĬãĤĬ
    0.17
     ØŃÙĬØ«
    0.16
     пÑĢиÑĩ
    0.16
    ýt
    0.16
    ØĽ
    0.15
     pues
    0.15
    ï¼ļ
    0.15
    Activations Density 0.050%

    No Known Activations