© 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 EvalsExportsinterp-engineNEWAPI 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. 11811
    Prev
    Next
    INDEX
    Explanations

    phrases ending with punctuation marks or special characters signifying excitement or emphasis

    oai_token-act-pair · gpt-3.5-turbo

    a specific character or symbol used repeatedly in conversation

    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
     Stuff
    -0.67
     Tanz
    -0.66
     Sahara
    -0.65
    iling
    -0.64
     cellphone
    -0.64
     Mobil
    -0.63
     Gaul
    -0.62
     Skydragon
    -0.62
     warnings
    -0.61
     Guant
    -0.61
    POSITIVE LOGITS
    should
    0.95
    felt
    0.94
    udes
    0.91
    ï¸ı
    0.90
    bec
    0.88
    shall
    0.84
    could
    0.84
    ved
    0.83
     deserves
    0.81
    ends
    0.81
    Activations Density 0.240%

    No Known Activations