© 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. 4-RES-JB
    11. 11468
    Prev
    Next
    INDEX
    Explanations

    instances where text mentions rewriting or rewritten content

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "rewrite" and its variations

    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
    ello
    -0.84
    aren
    -0.74
    ells
    -0.72
    amina
    -0.70
    ¬¼
    -0.70
    waukee
    -0.69
    ulhu
    -0.69
    Pa
    -0.69
    Hung
    -0.68
    Fi
    -0.68
    POSITIVE LOGITS
     rewritten
    1.20
     rewrite
    1.09
     rewriting
    0.94
    é»Ĵ
    0.75
     rethink
    0.73
     restruct
    0.72
     weave
    0.71
     versions
    0.70
    dated
    0.70
    UX
    0.66
    Activations Density 0.016%

    No Known Activations