© 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. 16508
    Prev
    Next
    INDEX
    Explanations

    instances where text is marked as part of a specific section or category, such as <*endoftext*> or news headings

    oai_token-act-pair · gpt-3.5-turbo

    the presence of empty tokens or line breaks in a text

    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
     tradem
    -0.75
     reper
    -0.73
    ilty
    -0.72
     amnesty
    -0.71
     eleph
    -0.69
     estranged
    -0.68
    thodox
    -0.68
     exting
    -0.66
    atively
    -0.66
    anamo
    -0.65
    POSITIVE LOGITS
    CLOSE
    1.13
    Untitled
    1.12
    WASHINGTON
    1.12
    Description
    1.10
    Abstract
    1.09
    Still
    1.06
    Trivia
    1.05
    Overview
    1.05
    CTV
    1.04
    Breaking
    1.02
    Activations Density 0.107%

    No Known Activations