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. 15409
    Prev
    Next
    INDEX
    Explanations

    phrases starting with "Not only" followed by a comparison or list of additional information

    oai_token-act-pair · gpt-3.5-turbo

    phrases emphasizing exclusivity or unique features

    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
    »Ĵ
    -0.68
    EStream
    -0.67
    ller
    -0.64
    iren
    -0.61
    NEY
    -0.60
    ashington
    -0.60
    ij士
    -0.59
    frac
    -0.59
    Īè
    -0.58
    ="/
    -0.57
    POSITIVE LOGITS
    tem
    0.64
    oped
    0.61
    onse
    0.59
    akia
    0.59
    ional
    0.58
    theless
    0.58
    asio
    0.57
    ivity
    0.57
    onia
    0.57
     physically
    0.56
    Activations Density 0.021%

    No Known Activations