© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 3-RES-JB
    11. 5104
    Prev
    Next
    INDEX
    Explanations

    proper nouns specifically related to the name "Pierce."

    oai_token-act-pair · gpt-3.5-turbo

    instances of the name "Pierce" and related references within a specific context

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    orical
    -0.70
    posted
    -0.68
    ĻĤ
    -0.68
    trump
    -0.66
    DonaldTrump
    -0.64
     looted
    -0.64
    Redditor
    -0.64
    ended
    -0.64
    ¥µ
    -0.63
    adapt
    -0.62
    POSITIVE LOGITS
     Bros
    1.08
     Pierce
    0.95
    burgh
    0.92
     Fulton
    0.85
     Pearce
    0.82
    ville
    0.81
    creen
    0.78
     Borough
    0.77
     Saunders
    0.76
     Bridges
    0.74
    Activations Density 0.011%

    No Known Activations