© 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 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. 4-RES-JB
    11. 1431
    Prev
    Next
    INDEX
    Explanations

    specific mentions of the term "Cersei," potentially related to a character or entity named Cersei

    oai_token-act-pair · gpt-3.5-turbo

    references to the word "Cer" and its variations, especially in contexts related to geography and specific entities

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

    No Comments

    Negative Logits
    ãĥĥãĥĪ
    -0.76
    edit
    -0.71
    sight
    -0.70
     Buffett
    -0.70
     Barkley
    -0.68
    lihood
    -0.66
    ãģį
    -0.65
     Psychiatric
    -0.64
    achusetts
    -0.64
    draft
    -0.62
    POSITIVE LOGITS
    idian
    0.96
    amic
    0.92
    isine
    0.89
    ategory
    0.87
    berus
    0.86
    prus
    0.85
    chio
    0.83
    acies
    0.83
    oder
    0.83
    osa
    0.83
    Activations Density 0.007%

    No Known Activations