© 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. 4-RES-JB
    11. 16814
    Prev
    Next
    INDEX
    Explanations

    mentions of mud with varying levels of activation, possibly indicating a focus on mud-related concepts or imagery

    oai_token-act-pair · gpt-3.5-turbo

    mentions of "mud."

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

    No Comments

    Negative Logits
    ECT
    -0.78
    ASED
    -0.75
    OHN
    -0.72
    OPLE
    -0.72
    ATURE
    -0.71
    Sax
    -0.69
    zona
    -0.67
    士
    -0.67
    ONT
    -0.65
     Stras
    -0.65
    POSITIVE LOGITS
    bath
    0.98
    bag
    0.90
    mist
    0.85
    geon
    0.85
    alore
    0.84
    weed
    0.83
     mud
    0.81
    hog
    0.80
    bags
    0.78
    iquette
    0.77
    Activations Density 0.010%

    No Known Activations