© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 5094
    Prev
    Next
    INDEX
    Explanations

    references to the platform Patreon and discussion related to funding or support on this platform

    oai_token-act-pair · gpt-3.5-turbo

    references to Patreon and discussions surrounding its role and issues related to morphine

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Negative Logits
     shooters
    -0.87
    Vision
    -0.70
     Shoot
    -0.68
     shooter
    -0.68
     Country
    -0.64
    Custom
    -0.64
     Combo
    -0.63
    cks
    -0.63
    sand
    -0.62
     Kids
    -0.62
    POSITIVE LOGITS
     Patreon
    2.37
     morphine
    1.91
     sympathetic
    1.45
     opium
    1.28
     Eliot
    1.22
     leptin
    1.08
    oxy
    1.06
     Thiel
    1.02
     gh
    1.01
     Yiannopoulos
    0.95
    Activations Density 0.053%

    No Known Activations