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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 3-RES-JB
    6. 19588
    Prev
    Next
    INDEX
    Explanations

    words related to the Arabic language or culture

    oai_token-act-pair · gpt-3.5-turbo

    mentions of the Arabic language and related terms

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ertodd
    -0.99
    alling
    -0.80
    pload
    -0.79
    ablishment
    -0.79
    kj
    -0.79
    vg
    -0.77
    lv
    -0.75
    hov
    -0.75
    lessly
    -0.74
     Wrestle
    -0.74
    POSITIVE LOGITS
     Arabic
    0.90
     accents
    0.85
     Hebrew
    0.80
    iyah
    0.79
    س
    0.79
     lingu
    0.79
     Aram
    0.78
     numer
    0.78
     transl
    0.78
     translations
    0.77
    Activations Density 0.005%

    No Known Activations