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. 23553
    Prev
    Next
    INDEX
    Explanations

    events related to hostage situations and their resolutions

    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
    emax
    -0.17
    oucher
    -0.15
    EMON
    -0.15
    اØŃÛĮ
    -0.14
    kyt
    -0.14
    /Foundation
    -0.14
     sched
    -0.14
    éŀ
    -0.13
    iaux
    -0.13
    å¹²
    -0.13
    POSITIVE LOGITS
     hostage
    0.42
     hostages
    0.39
     standoff
    0.32
     Negot
    0.30
     negot
    0.29
     negotiation
    0.29
     negotiations
    0.27
     siege
    0.26
     barric
    0.26
     situation
    0.24
    Activations Density 0.058%

    No Known Activations