© 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. 3-RES-JB
    11. 17451
    Prev
    Next
    INDEX
    Explanations

    phrases indicating a comparison or contrast

    oai_token-act-pair · gpt-3.5-turbo

    phrases that contain the word "which."

    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
    Behind
    -0.85
    athi
    -0.77
    Bas
    -0.74
    hat
    -0.66
    WARE
    -0.63
    grim
    -0.63
    Gy
    -0.63
    BOX
    -0.61
    GY
    -0.60
    ben
    -0.60
    POSITIVE LOGITS
    soever
    0.96
     contrasts
    0.77
    xual
    0.76
     resulted
    0.74
     consists
    0.72
     comprises
    0.71
     incidentally
    0.70
     brings
    0.69
     corresponds
    0.69
     begs
    0.68
    Activations Density 0.135%

    No Known Activations