© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 17243
    Prev
    Next
    INDEX
    Explanations

    adjectives related to being easily remembered or having a lasting impact

    oai_token-act-pair · gpt-3.5-turbo

    references to memorable or unforgettable moments and experiences

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Negative Logits
    ©¶æ
    -0.77
    sterdam
    -0.76
    chen
    -0.73
    soon
    -0.69
    chens
    -0.67
    upon
    -0.66
    á
    -0.66
    ilon
    -0.66
    raged
    -0.65
    hm
    -0.65
    POSITIVE LOGITS
     memorable
    1.07
     unforgettable
    0.93
     memories
    0.87
     moments
    0.86
    isable
    0.76
    wcs
    0.76
     remem
    0.75
    theless
    0.74
     cinematic
    0.72
    TextColor
    0.69
    Activations Density 0.011%

    No Known Activations