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. 31385
    Prev
    Next
    INDEX
    Explanations

    references to animal welfare and the evaluation of conditions or qualities

    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
    .scalablytyped
    -0.21
    ozor
    -0.18
    .GetResponse
    -0.17
    _PHP
    -0.15
    .MaxLength
    -0.15
    alat
    -0.14
    heel
    -0.14
     конÑĤÑĢолÑĮ
    -0.14
    omez
    -0.14
    .Ultra
    -0.14
    POSITIVE LOGITS
     fit
    0.53
     worthy
    0.47
     worth
    0.43
     suitable
    0.42
    suit
    0.42
     Fit
    0.39
    Fit
    0.39
    -fit
    0.38
    fit
    0.37
    worthy
    0.37
    Activations Density 0.027%

    No Known Activations