© 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. 0-RES-JB
    11. 12348
    Prev
    Next
    INDEX
    Explanations

    the abbreviation "pg" followed by a numerical value, potentially related to pages or pagination

    oai_token-act-pair · gpt-3.5-turbo

    instances of the term "pg" that likely relate to pagination in technical or programming 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
     unmanned
    -0.64
     bona
    -0.62
     Scar
    -0.62
     disrespect
    -0.61
     recal
    -0.61
     fixed
    -0.58
     Belt
    -0.58
     Wand
    -0.58
     interrupt
    -0.57
    eer
    -0.57
    POSITIVE LOGITS
    pg
    4.84
     pg
    1.85
    PG
    1.76
    Pg
    1.41
    vg
    1.29
    peg
    1.24
    RPG
    1.19
    gp
    1.17
    pb
    1.08
    pc
    1.08
    Activations Density 0.013%

    No Known Activations