© 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. 9073
    Prev
    Next
    INDEX
    Explanations

    instances of the word "new" and its variations in various contexts

    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
     newfound
    -0.19
     нового
    -0.15
     newly
    -0.15
    .central
    -0.15
     newest
    -0.15
    ITES
    -0.15
    Äįel
    -0.15
    uate
    -0.14
    골
    -0.14
    phalt
    -0.14
    POSITIVE LOGITS
    bies
    0.31
    sworth
    0.27
    -found
    0.24
    sw
    0.23
    swire
    0.23
     Zealand
    0.23
    é²ľ
    0.23
    sg
    0.22
    -old
    0.21
    ish
    0.21
    Activations Density 0.101%

    No Known Activations