Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 23699
    Prev
    Next
    INDEX
    Explanations

    past participles and past tense verbs

    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
    лаÑģÑĤи
    -0.16
    èŃĺ
    -0.15
    Äįet
    -0.14
    inyin
    -0.14
    zelf
    -0.14
    ward
    -0.14
    à¸Ľà¸£à¸°à¸¡à¸²à¸ĵ
    -0.14
    _procs
    -0.14
    ing
    -0.14
    bild
    -0.13
    POSITIVE LOGITS
     up
    0.30
     away
    0.29
     out
    0.28
     down
    0.28
     into
    0.27
    -up
    0.27
    -down
    0.25
     off
    0.25
    -out
    0.24
    -off
    0.22
    Activations Density 0.027%

    No Known Activations