© 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. 8-RES-JB
    11. 3245
    Prev
    Next
    INDEX
    Explanations

    mentions of the word "Pad" either in isolation or as part of a longer word

    oai_token-act-pair · gpt-3.5-turbo

    references to the term "pad" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Heller
    -0.67
    eger
    -0.63
    çİĭ
    -0.61
    naires
    -0.61
    IDER
    -0.60
     Samar
    -0.59
    ews
    -0.58
     Ultron
    -0.57
    ously
    -0.57
     Coh
    -0.57
    POSITIVE LOGITS
    awan
    1.30
    lock
    1.04
    locks
    0.96
    het
    0.95
    locked
    0.94
    rice
    0.89
    asus
    0.87
    pole
    0.85
    geon
    0.84
    agon
    0.84
    Activations Density 0.026%

    No Known Activations