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. 21346
    Prev
    Next
    INDEX
    Explanations

    references to or mentions of the word "Bunny"

    oai_token-act-pair · gpt-3.5-turbo

    references to "Bunny" and related terms, particularly in a whimsical or light-hearted context

    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
    umar
    -0.76
    ERA
    -0.72
    ESCO
    -0.69
     oxide
    -0.69
    ridor
    -0.67
    olith
    -0.66
     SATA
    -0.66
    itage
    -0.65
    eneg
    -0.65
    ignment
    -0.64
    POSITIVE LOGITS
     Bunny
    3.56
     bunny
    1.90
    liness
    1.40
     Wilde
    1.27
     imperson
    1.12
    vernment
    1.10
    nard
    1.05
    unny
    0.99
     droid
    0.95
     fox
    0.94
    Activations Density 0.057%

    No Known Activations