© 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. 41664
    Prev
    Next
    INDEX
    Explanations

    phrases related to adjustments or alterations in policies or conditions

    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
    hots
    -0.20
    hart
    -0.20
    ho
    -0.19
    aug
    -0.19
    iest
    -0.19
    hire
    -0.19
    hunt
    -0.19
    ily
    -0.19
    hand
    -0.18
    hog
    -0.18
    POSITIVE LOGITS
    ting
    1.05
    ters
    0.69
    ted
    0.64
    tings
    0.62
    TING
    0.62
    ter
    0.59
    te
    0.52
    ten
    0.49
    tery
    0.48
    ty
    0.48
    Activations Density 0.030%

    No Known Activations