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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    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. 29878
    Prev
    Next
    INDEX
    Explanations

    phrases referencing official statements or announcements

    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
    Operand
    -0.15
    inosaur
    -0.15
     Problems
    -0.14
    .eql
    -0.14
    ondheim
    -0.14
    keley
    -0.14
    cisi
    -0.14
    ãĥ¼ãĥķ
    -0.13
    mal
    -0.13
    ank
    -0.13
    POSITIVE LOGITS
     matter
    0.63
    matter
    0.48
     Matter
    0.44
     situation
    0.38
     matters
    0.33
     story
    0.30
     Matters
    0.28
    story
    0.27
     mater
    0.26
     subject
    0.25
    Activations Density 0.036%

    No Known Activations