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. 21983
    Prev
    Next
    INDEX
    Explanations

    names of historical figures, most notably "Borgia" and "Mandela"

    oai_token-act-pair · gpt-3.5-turbo

    proper nouns and names, particularly related to individuals and fictional characters

    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
    hire
    -0.84
     SCHOOL
    -0.76
     Egypt
    -0.69
    school
    -0.66
    yrim
    -0.66
    yrus
    -0.65
     scheme
    -0.65
     sho
    -0.65
     Safari
    -0.63
     Schools
    -0.63
    POSITIVE LOGITS
     Borg
    4.21
     Berman
    1.36
     Berger
    1.21
    borg
    1.15
     Klingon
    1.09
     Bj
    1.04
     Cald
    1.02
     Ortiz
    1.00
     Vec
    1.00
     Voyager
    0.98
    Activations Density 0.029%

    No Known Activations