© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 2316
    Prev
    Next
    INDEX
    Explanations

    mentions of things being unexciting or subpar, specifically using the term "lame."

    oai_token-act-pair · gpt-3.5-turbo

    negative descriptors, particularly "lame" and specific references to "canned" and "mediocre" content

    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
    Ul
    -0.77
    verty
    -0.75
     Liberties
    -0.75
    roud
    -0.74
     Sovere
    -0.73
    rawl
    -0.70
     Swap
    -0.69
    hatt
    -0.68
    Merit
    -0.68
    bits
    -0.67
    POSITIVE LOGITS
     lame
    2.41
     canned
    2.12
     medi
    1.80
     bottled
    1.22
     CGI
    1.16
    SON
    1.07
    emo
    0.98
     orchestr
    0.97
     scripted
    0.89
     mediated
    0.88
    Activations Density 0.052%

    No Known Activations