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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 21801
    Prev
    Next
    INDEX
    Explanations

    the concept of "manner" or the way in which actions are performed

    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
    y
    -0.20
    sel
    -0.19
    sy
    -0.17
    sg
    -0.16
    airs
    -0.16
    oted
    -0.15
    noc
    -0.15
    yd
    -0.15
    young
    -0.15
    nic
    -0.15
    POSITIVE LOGITS
    isms
    0.40
    ly
    0.26
    ed
    0.24
    Ø©
    0.23
    íģ¼
    0.22
    ism
    0.20
    lust
    0.19
    estro
    0.19
    istic
    0.19
    ifest
    0.18
    Activations Density 0.015%

    No Known Activations