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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 54389
    Prev
    Next
    INDEX
    Explanations

    negations and phrases that express opposition or disagreement

    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
    lan
    -0.15
    poke
    -0.15
    andes
    -0.15
    claimer
    -0.15
    LAN
    -0.14
    æŦ
    -0.14
    ingers
    -0.14
    jom
    -0.14
     /^\
    -0.14
     thôi
    -0.14
    POSITIVE LOGITS
     rocket
    0.31
     about
    0.22
     Rocket
    0.22
    rocket
    0.21
     complicated
    0.21
     negot
    0.19
     mutually
    0.18
     optional
    0.18
     rockets
    0.18
     only
    0.18
    Activations Density 0.043%

    No Known Activations