© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 63895
    Prev
    Next
    INDEX
    Explanations

    connective phrases indicating causation or condition

    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
    /her
    -0.14
    asi
    -0.14
    adlo
    -0.14
    agli
    -0.13
     ourselves
    -0.13
    indi
    -0.13
    ully
    -0.13
    该
    -0.13
    á»ĥ
    -0.13
    cludes
    -0.12
    POSITIVE LOGITS
     he
    0.30
     she
    0.26
     they
    0.26
     He
    0.22
    ä»ĸ
    0.19
     They
    0.18
     She
    0.18
     он
    0.18
     ä»ĸ
    0.17
    she
    0.17
    Activations Density 0.417%

    No Known Activations