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. 3568
    Prev
    Next
    INDEX
    Explanations

    references to a supernatural entity or concept called "Spirit."

    oai_token-act-pair · gpt-3.5-turbo

    references to the term "Spirit"

    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
    ELY
    -0.77
    cro
    -0.75
    00007
    -0.70
     clamp
    -0.68
    zai
    -0.68
    azine
    -0.68
    define
    -0.66
     hashes
    -0.65
     bump
    -0.64
     workaround
    -0.64
    POSITIVE LOGITS
     Spirit
    3.89
    Spirit
    2.77
     Spirits
    2.23
    spirit
    2.05
     spirit
    1.99
     Ghost
    1.63
     spirits
    1.49
     Spiritual
    1.47
     Soul
    1.40
     Myst
    1.24
    Activations Density 0.013%

    No Known Activations