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. 8-RES-JB
    6. 12066
    Prev
    Next
    INDEX
    Explanations

    instances of the word "First" as the first item in a list or sequence, or when referring to a primary point or step

    oai_token-act-pair · gpt-3.5-turbo

    references to sequential reasoning or itemized lists

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    sed
    -0.72
    "},"
    -0.66
    phe
    -0.61
    GA
    -0.59
     passer
    -0.58
     +++
    -0.58
    walking
    -0.57
    gone
    -0.55
    sung
    -0.55
     Burning
    -0.55
    POSITIVE LOGITS
     involves
    0.69
     TTL
    0.69
     Firstly
    0.68
     namely
    0.66
     Ones
    0.66
     relates
    0.65
     awa
    0.65
    ãĤ´ãĥ³
    0.64
     arose
    0.63
    %:
    0.63
    Activations Density 0.092%

    No Known Activations