© 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. 15764
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "has" and other forms indicating possession or existence

    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
     personally
    -0.20
    aight
    -0.17
     himself
    -0.16
    aeda
    -0.16
    roleum
    -0.15
    aft
    -0.15
    ead
    -0.15
    Enumer
    -0.15
    аÑĢÑı
    -0.15
     personal
    -0.14
    POSITIVE LOGITS
     itself
    0.51
     its
    0.34
     Its
    0.29
    Its
    0.27
    its
    0.19
    åħ¶
    0.18
    ï¼Įå®ĥ
    0.17
     derec
    0.17
     consisted
    0.15
    å®ĥ
    0.15
    Activations Density 0.053%

    No Known Activations