Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 23631
    Prev
    Next
    INDEX
    Explanations

    references to specific publications and their contributors

    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
    ibs
    -0.18
     Manhattan
    -0.17
     NYC
    -0.16
    .NewReader
    -0.15
    >NN
    -0.15
     Harlem
    -0.15
    adiens
    -0.15
    ensi
    -0.14
    ätt
    -0.14
    "user
    -0.14
    POSITIVE LOGITS
     Washington
    0.34
     washington
    0.33
    Washington
    0.30
     DC
    0.28
    DC
    0.28
     dc
    0.27
     Redskins
    0.27
    ashington
    0.26
    ASHINGTON
    0.24
     Wa
    0.23
    Activations Density 0.016%

    No Known Activations