© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 15341
    Prev
    Next
    INDEX
    Explanations

    instances of the word "haha" at varying intensities

    oai_token-act-pair · gpt-3.5-turbo

    expressions of laughter or amusement

    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
    lining
    -0.86
    lines
    -0.82
    master
    -0.77
    lest
    -0.73
     directions
    -0.73
    ership
    -0.69
     direction
    -0.68
    picking
    -0.67
    craft
    -0.66
    writing
    -0.66
    POSITIVE LOGITS
    ahah
    1.12
    aha
    1.07
    awk
    0.97
    ibaba
    0.97
    awks
    0.90
    HAHA
    0.89
    uthor
    0.88
    qua
    0.88
    ghan
    0.87
    UGH
    0.85
    Activations Density 0.014%

    No Known Activations