© 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. 8-RES-JB
    11. 10676
    Prev
    Next
    INDEX
    Explanations

    phrases related to dramatic emphasis or tension

    oai_token-act-pair · gpt-3.5-turbo

    punctuation and its emotional emphasis in text

    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
     rall
    -0.88
     encount
    -0.75
     valued
    -0.75
     volunte
    -0.74
    perty
    -0.73
     necess
    -0.72
     surpr
    -0.72
    inement
    -0.71
     withd
    -0.70
     threatened
    -0.69
    POSITIVE LOGITS
     Anyway
    1.28
     Yeah
    1.14
     Exactly
    1.12
     Didn
    1.12
     Sorry
    1.10
     Alright
    1.09
     Okay
    1.09
     Thanks
    1.06
     Absolutely
    1.05
     Lots
    1.04
    Activations Density 0.193%

    No Known Activations