© 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. 40940
    Prev
    Next
    INDEX
    Explanations

    references to military exercises and cooperation between countries

    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
    ÃĹ↵↵
    -0.15
    egas
    -0.14
    ìĿ´ìĬĪ
    -0.14
    ofilm
    -0.14
    oven
    -0.14
    SWEP
    -0.13
    emoc
    -0.13
    olla
    -0.13
    .esp
    -0.13
     ESP
    -0.13
    POSITIVE LOGITS
     exercise
    0.59
     Exercise
    0.57
     exercises
    0.57
    Exercise
    0.51
     Exercises
    0.49
    exercise
    0.48
     exerc
    0.44
     exercising
    0.43
     drills
    0.42
     exercised
    0.42
    Activations Density 0.040%

    No Known Activations