© 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. 3-RES-JB
    11. 4964
    Prev
    Next
    INDEX
    Explanations

    phrases related to individuals with the surname "Steele" and "Woodward"

    oai_token-act-pair · gpt-3.5-turbo

    mentions of specific individuals, particularly Steele and Woodward

    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
    isp
    -0.79
    ahime
    -0.79
    alg
    -0.78
     [|
    -0.73
    outh
    -0.71
    uilt
    -0.70
    istant
    -0.68
    udeb
    -0.66
    ariat
    -0.66
    ä¸ī
    -0.65
    POSITIVE LOGITS
     Steele
    1.28
     dossier
    0.92
    lements
    0.84
    lla
    0.82
    rence
    0.76
    llers
    0.75
    burgh
    0.74
     Township
    0.73
    sburg
    0.73
    ling
    0.72
    Activations Density 0.012%

    No Known Activations