© 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. 0-RES-JB
    11. 12600
    Prev
    Next
    INDEX
    Explanations

    words or phrases written in a specific non-English script

    oai_token-act-pair · gpt-3.5-turbo

    specific non-Latin characters or symbols, possibly related to a particular language or script

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     scroll
    -0.72
     Rush
    -0.71
     Oracle
    -0.70
     Riley
    -0.68
     routine
    -0.65
     Wa
    -0.65
     substitute
    -0.64
     companion
    -0.63
     James
    -0.63
     Rider
    -0.63
    POSITIVE LOGITS
    Ñ
    3.97
    е
    2.51
    Ñı
    2.38
    ÑĤ
    2.36
    о
    2.28
    Ð
    2.23
    д
    2.23
    к
    2.17
    Ñģ
    2.13
    м
    2.13
    Activations Density 0.009%

    No Known Activations