© 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
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 54752
    Prev
    Next
    INDEX
    Explanations

    phrases related to legal violations and the consequences of disregarding rules

    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
    ÎŃÏģα
    -0.17
    ãģ¤ãģ¶
    -0.15
    ãģŁãģı
    -0.15
    rror
    -0.15
     ÙĦÙĥÙĦ
    -0.14
    addtogroup
    -0.14
    豪
    -0.14
    _mapper
    -0.13
    orgot
    -0.13
    زÙĩ
    -0.13
    POSITIVE LOGITS
     established
    0.29
     existing
    0.24
     rules
    0.24
     norms
    0.21
    -established
    0.21
     protocol
    0.21
     accepted
    0.21
     Established
    0.20
     the
    0.20
     regulations
    0.19
    Activations Density 0.209%

    No Known Activations