© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 16868
    Prev
    Next
    INDEX
    Explanations

    references to "SEE" in the text

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "see" and related phrases, indicating references to visual content or citations

    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
     Bord
    -0.74
     Parm
    -0.67
    agh
    -0.65
    fman
    -0.64
    odan
    -0.61
    ender
    -0.61
    aration
    -0.61
     Journals
    -0.61
     Goldstein
    -0.61
     Ascension
    -0.61
    POSITIVE LOGITS
    SEE
    4.72
     SEE
    1.51
    DON
    1.22
    SPONSORED
    1.10
    LI
    1.01
    QUI
    0.97
    CHECK
    0.90
    TRY
    0.86
    EMA
    0.86
    YES
    0.85
    Activations Density 0.032%

    No Known Activations