© 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 EvalsExportsinterp-engineNEWAPI 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. 3-RES-JB
    11. 16852
    Prev
    Next
    INDEX
    Explanations

    content-related text often found at the beginning or end of an article or document

    oai_token-act-pair · gpt-3.5-turbo

    references to sections or categories in lists or documents

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    edIn
    -0.80
     pens
    -0.73
    emouth
    -0.71
     occas
    -0.70
    wagen
    -0.70
     rooft
    -0.69
    nesday
    -0.67
     entitle
    -0.67
     scraps
    -0.66
     therap
    -0.65
    POSITIVE LOGITS
    ?]
    1.37
    ËĪ
    1.34
     edit
    1.24
    !]
    1.14
    actionDate
    1.12
    sic
    1.10
    emphasis
    1.06
    Footnote
    1.03
    %]
    1.01
    Pg
    1.00
    Activations Density 0.041%

    No Known Activations