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. 18864
    Prev
    Next
    INDEX
    Explanations

    terms related to specific words, such as "Yamuna," "Yamazaki," and "Shimano."

    oai_token-act-pair · gpt-3.5-turbo

    references to the name "Yam" or variants of it

    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
    ttes
    -0.72
    byter
    -0.69
     parity
    -0.68
     Whitman
    -0.65
     commissioners
    -0.65
     Purg
    -0.64
     cavity
    -0.64
     creep
    -0.61
     AAP
    -0.61
     miss
    -0.58
    POSITIVE LOGITS
    azaki
    1.23
    amoto
    1.21
    atsu
    1.05
    amaz
    1.03
    aku
    1.03
    aji
    1.02
     Yam
    1.00
    oka
    0.96
    asaki
    0.96
    assian
    0.93
    Activations Density 0.006%

    No Known Activations