© 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. 1-RES-JB
    11. 2263
    Prev
    Next
    INDEX
    Explanations

    words related to injuries

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "wound" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    alty
    -0.74
    acons
    -0.72
    anooga
    -0.69
    sol
    -0.67
    effic
    -0.67
     pudding
    -0.66
     practicable
    -0.65
    agically
    -0.64
    outh
    -0.63
    extra
    -0.62
    POSITIVE LOGITS
     wound
    1.16
     wounds
    0.94
    robe
    0.82
     Whedon
    0.82
    terson
    0.76
    FontSize
    0.73
    puter
    0.72
    bow
    0.71
     4090
    0.69
    bolt
    0.69
    Activations Density 0.011%

    No Known Activations