© 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. 3-RES-JB
    11. 3253
    Prev
    Next
    INDEX
    Explanations

    mentions of physical cleanliness or dirtiness

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "dirty" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    yip
    -0.89
    HCR
    -0.86
    ensical
    -0.85
    zai
    -0.84
    icultural
    -0.84
    */(
    -0.82
    ajor
    -0.82
    ommel
    -0.82
    iphate
    -0.81
    uther
    -0.80
    POSITIVE LOGITS
     laundry
    0.97
     dirty
    0.90
     linen
    0.89
     cleaner
    0.82
     diaper
    0.80
     diapers
    0.79
     mole
    0.78
    dirty
    0.76
     stains
    0.73
    clean
    0.72
    Activations Density 0.010%

    No Known Activations