© 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. 1026
    Prev
    Next
    INDEX
    Explanations

    phrases related to delaying or putting something on hold

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "hold."

    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
    icycle
    -0.66
    xon
    -0.63
     Audit
    -0.63
    cussion
    -0.61
    eatured
    -0.60
    endez
    -0.60
    incial
    -0.59
    TeX
    -0.59
     Murder
    -0.59
    ppard
    -0.58
    POSITIVE LOGITS
    hold
    1.45
    holding
    0.91
    erness
    0.87
     hold
    0.84
    holder
    0.83
    backs
    0.82
    ings
    0.81
    holders
    0.80
    pring
    0.79
    back
    0.78
    Activations Density 0.019%

    No Known Activations