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. 13784
    Prev
    Next
    INDEX
    Explanations

    sequences of repeated special characters

    oai_token-act-pair · gpt-3.5-turbo

    patterns of repeated characters or symbols

    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
     Samar
    -0.78
    achus
    -0.71
    ake
    -0.62
    milo
    -0.62
     denying
    -0.62
     Watkins
    -0.60
    adelphia
    -0.59
     minds
    -0.59
     pegged
    -0.59
    anke
    -0.58
    POSITIVE LOGITS
    --------------------------------------------------------
    1.03
    --------------------------------
    0.89
    =-
    0.87
    ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
    0.86
    EGIN
    0.80
    =-=-=-=-=-=-=-=-
    0.77
    ++++
    0.77
    ----
    0.77
    âĸł
    0.77
    ---
    0.76
    Activations Density 0.342%

    No Known Activations