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. 3164
    Prev
    Next
    INDEX
    Explanations

    proper nouns, specifically names like "Williams," "Carter-Williams," and "BioWare."

    oai_token-act-pair · gpt-3.5-turbo

    mentions of the name "Williams" in various contexts

    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
     Gener
    -0.77
     Ord
    -0.73
     Dest
    -0.72
     unic
    -0.70
     Uri
    -0.70
     Mandatory
    -0.69
     Arri
    -0.69
     Vac
    -0.66
     dra
    -0.65
     Def
    -0.65
    POSITIVE LOGITS
    Williams
    2.94
    Roberts
    1.52
    Adams
    1.29
    Bride
    1.24
    Ware
    1.22
    Moore
    1.14
    zyk
    1.13
    DNA
    1.13
    William
    1.10
    Harris
    1.10
    Activations Density 0.045%

    No Known Activations