© 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
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 42213
    Prev
    Next
    INDEX
    Explanations

    phrases related to legal agreements and arbitration processes

    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
    umo
    -0.15
    SWG
    -0.15
     PlzeÅĪ
    -0.15
    à¸ĩà¸Ĥ
    -0.15
    à¸Ĺาà¸Ļ
    -0.15
    aghan
    -0.14
     erotisk
    -0.14
     cerv
    -0.14
     cerco
    -0.14
    Prostit
    -0.14
    POSITIVE LOGITS
     arbitration
    0.46
     Arbit
    0.46
     arbit
    0.43
     Arb
    0.41
     arbitr
    0.41
     arb
    0.40
     dispute
    0.39
    仲
    0.39
     Med
    0.39
    -ar
    0.38
    Activations Density 0.027%

    No Known Activations