© 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. 24504
    Prev
    Next
    INDEX
    Explanations

    phrases related to evasion or getting away with something

    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
    iano
    -0.14
    ino
    -0.14
    arin
    -0.14
    akens
    -0.14
    handleRequest
    -0.14
    Ùģات
    -0.14
     Brend
    -0.13
    ä¼
    -0.13
    loid
    -0.13
    _REGS
    -0.13
    POSITIVE LOGITS
     away
    0.65
    away
    0.55
     Away
    0.54
    Away
    0.48
    -away
    0.44
    aways
    0.29
     getaway
    0.27
     weg
    0.26
     get
    0.25
     Get
    0.24
    Activations Density 0.018%

    No Known Activations