© 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. 3-RES-JB
    11. 3238
    Prev
    Next
    INDEX
    Explanations

    the abbreviation "MK" followed by a number or Roman numeral designation

    oai_token-act-pair · gpt-3.5-turbo

    references to a specific comic book series or its associated characters

    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
    Interstitial
    -0.73
    ================================================================
    -0.72
     Reaper
    -0.72
    aldo
    -0.68
    istics
    -0.68
    ãĥ¼ãĥ³
    -0.68
    ãĥ¼ãĥĨ
    -0.67
    ãĥ³ãĤ¸
    -0.64
    cean
    -0.64
    ordon
    -0.64
    POSITIVE LOGITS
    MK
    0.99
     MK
    0.93
    Ultra
    0.87
    enzie
    0.82
    JV
    0.80
    ULT
    0.78
    ellen
    0.77
    adish
    0.77
     Ultra
    0.75
    intosh
    0.74
    Activations Density 0.011%

    No Known Activations