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. 0-RES-JB
    6. 3096
    Prev
    Next
    INDEX
    Explanations

    a specific string denoting the end of a text or message

    oai_token-act-pair · gpt-3.5-turbo

    symbols or characters typically used to denote navigation or formatting in digital content

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ��
    -0.66
    witz
    -0.66
    ����
    -0.64
    nell
    -0.63
    ubi
    -0.62
    daq
    -0.61
    ,—
    -0.57
     prost
    -0.56
     Inquis
    -0.56
     Carlton
    -0.56
    POSITIVE LOGITS
     >
    3.65
     >>
    2.32
     ><
    2.28
    >
    2.22
    >.
    1.86
     >=
    1.85
     âī¥
    1.85
    >,
    1.83
     >>>
    1.79
    _>
    1.76
    Activations Density 0.011%

    No Known Activations