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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 25645
    Prev
    Next
    INDEX
    Explanations

    terms related to data input and processing in a technical context

    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
    ettle
    -0.15
     Output
    -0.15
    ita
    -0.14
    aho
    -0.14
    ITA
    -0.14
    awa
    -0.14
    (Output
    -0.14
    incer
    -0.13
    ExecutionContext
    -0.13
    ụ
    -0.13
    POSITIVE LOGITS
     entry
    0.93
     enter
    0.83
     Entry
    0.80
    entry
    0.77
    -entry
    0.75
     enters
    0.75
     entered
    0.74
     Enter
    0.74
     entries
    0.73
    Entry
    0.72
    Activations Density 0.028%

    No Known Activations