© 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. 1-RES-JB
    11. 18663
    Prev
    Next
    INDEX
    Explanations

    words related to online presence and communication such as usernames, emails, and website addresses

    oai_token-act-pair · gpt-3.5-turbo

    the acronym "TH" and its various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ãĤŃ
    -0.81
    atories
    -0.69
    ãĤ±
    -0.68
    mens
    -0.67
    archives
    -0.67
    ãĤ°
    -0.67
    unct
    -0.66
    olitan
    -0.66
    ãĥ¯
    -0.66
    dep
    -0.64
    POSITIVE LOGITS
    ulhu
    1.05
    TH
    1.01
    IRD
    0.98
    ERE
    0.93
    irteen
    0.92
    orne
    0.89
    ursday
    0.87
    DAY
    0.85
    urst
    0.85
    OUGH
    0.85
    Activations Density 0.006%

    No Known Activations