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. 21751
    Prev
    Next
    INDEX
    Explanations

    references to technical aspects of nodes or node-related concepts in a computational context

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "node" and its variations, indicating a focus on network or data structure terminology

    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
     Guinness
    -0.83
    amaz
    -0.77
     MGM
    -0.74
    hol
    -0.73
     Ghostbusters
    -0.72
    exempt
    -0.72
    ritz
    -0.72
     Franks
    -0.71
     Hogan
    -0.69
     Pepsi
    -0.68
    POSITIVE LOGITS
     node
    3.62
    node
    2.94
     Node
    2.90
     nodes
    2.77
    Node
    2.48
     npm
    1.47
    odes
    1.27
     vertex
    1.25
     cluster
    1.13
     tree
    1.11
    Activations Density 0.021%

    No Known Activations