© 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. 13945
    Prev
    Next
    INDEX
    Explanations

    words related to physical distress or discomfort, specifically related to "writhing."

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of words starting with "wr"

    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
     Trip
    -0.70
     Bars
    -0.66
     Bitcoins
    -0.65
     Locations
    -0.64
    mits
    -0.64
     Grounds
    -0.63
    ultural
    -0.63
    bed
    -0.62
     Places
    -0.60
    Beta
    -0.60
    POSITIVE LOGITS
     wr
    3.74
     Wr
    1.53
    wr
    1.46
     wrinkles
    1.46
     wrench
    1.14
     twist
    1.12
     scr
    1.09
     wor
    1.08
     tug
    1.07
     wrest
    1.02
    Activations Density 0.019%

    No Known Activations