© 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
    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. 18921
    Prev
    Next
    INDEX
    Explanations

    names of people or titles with an honorific like "Imam" or "Former board member."

    oai_token-act-pair · gpt-3.5-turbo

    proper nouns, particularly names

    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
    itate
    -0.68
    Ñı
    -0.64
     written
    -0.63
     read
    -0.61
     Dynam
    -0.61
     bothered
    -0.60
     correctly
    -0.60
    stem
    -0.59
     +/-
    -0.59
     under
    -0.58
    POSITIVE LOGITS
    maxwell
    0.94
    anamo
    0.83
     Seym
    0.78
    emort
    0.77
    iferation
    0.76
     advoc
    0.76
     caut
    0.75
    Downloadha
    0.75
    odan
    0.74
     eleph
    0.71
    Activations Density 0.000%

    No Known Activations