© 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. 14584
    Prev
    Next
    INDEX
    Explanations

    references to emails, often with mentions of specific numbers or actions related to them

    oai_token-act-pair · gpt-3.5-turbo

    references to emails and related correspondence

    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
     Edinburgh
    -0.74
    OV
    -0.73
     Gibraltar
    -0.71
     Glasgow
    -0.71
    iasis
    -0.70
    scope
    -0.70
    INAL
    -0.69
    axis
    -0.68
    CTV
    -0.67
     Helsinki
    -0.67
    POSITIVE LOGITS
     Emails
    1.10
     emails
    1.06
     inbox
    1.03
    ileaks
    0.95
     newsletters
    0.93
     messages
    0.91
     correspondence
    0.89
     exchanged
    0.89
     memos
    0.89
     dumps
    0.87
    Activations Density 0.028%

    No Known Activations