© 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. 4-RES-JB
    11. 15468
    Prev
    Next
    INDEX
    Explanations

    instances of non-compliance or failure to meet expectations

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "failing."

    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
    Revolution
    -0.74
    opter
    -0.69
    andise
    -0.68
    dress
    -0.67
    onen
    -0.67
    arger
    -0.66
    opot
    -0.66
    utra
    -0.66
    ization
    -0.65
     Extensions
    -0.64
    POSITIVE LOGITS
     miser
    1.29
    lect
    0.76
    afe
    0.75
     dism
    0.75
     horribly
    0.74
    fully
    0.72
     catast
    0.71
     failing
    0.70
     grades
    0.70
    ppe
    0.70
    Activations Density 0.013%

    No Known Activations