© 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. 0-RES-JB
    11. 15008
    Prev
    Next
    INDEX
    Explanations

    specific terms or concepts mentioned previously in the text

    oai_token-act-pair · gpt-3.5-turbo

    terms that suggest references to previously mentioned or notable entities and conditions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     renew
    -0.62
     Tasmania
    -0.62
    sted
    -0.61
     compatible
    -0.58
     âĢº
    -0.58
     divided
    -0.58
     cooper
    -0.56
     POLIT
    -0.56
    Scotland
    -0.56
    seys
    -0.56
    POSITIVE LOGITS
     aforementioned
    3.05
     dreaded
    1.95
     infamous
    1.89
     requisite
    1.39
     coveted
    1.34
     afore
    1.30
     proverbial
    1.26
     admittedly
    1.25
     pesky
    1.23
     latter
    1.22
    Activations Density 0.042%

    No Known Activations