© 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. 3-RES-JB
    11. 4765
    Prev
    Next
    INDEX
    Explanations

    um and proper nouns

    oai_token-act-pair · gpt-3.5-turbo

    filler words or interjections expressing hesitation

    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
     VIDEOS
    -0.80
    å£
    -0.78
    ships
    -0.73
    æ©
    -0.72
    advertisement
    -0.71
    Adds
    -0.70
    YouTube
    -0.67
    éĹĺ
    -0.63
     vents
    -0.62
    sburgh
    -0.62
    POSITIVE LOGITS
     Um
    1.00
    Um
    0.95
    uds
    0.92
    unia
    0.87
    iple
    0.84
    iatus
    0.84
    ajo
    0.83
    rha
    0.83
    rah
    0.82
    isl
    0.82
    Activations Density 0.010%

    No Known Activations