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

    Neuronpedia

    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. 4796
    Prev
    Next
    INDEX
    Explanations

    words related to violent actions and sticky substances

    oai_token-act-pair · gpt-3.5-turbo

    references to lynching and related concepts of difficulty or hold

    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
     Rack
    -0.89
     Goldberg
    -0.81
    yssey
    -0.74
     Bei
    -0.73
     Vale
    -0.72
    robe
    -0.72
    burgh
    -0.72
    idth
    -0.71
    aughs
    -0.68
     Rox
    -0.67
    POSITIVE LOGITS
     lyn
    2.00
     sticky
    1.71
     VAT
    1.45
     hel
    1.36
     glue
    1.32
     decom
    1.24
     READ
    1.14
     hanged
    1.10
     tor
    0.99
     ty
    0.99
    Activations Density 0.052%

    No Known Activations