© 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. 3-RES-JB
    11. 6115
    Prev
    Next
    INDEX
    Explanations

    common phrases or expressions used in discussions or written texts

    oai_token-act-pair · gpt-3.5-turbo

    references to various types or categories of behaviors, actions, or phenomena

    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
     spokeswoman
    -0.67
    \\\\\\\\
    -0.67
     Courier
    -0.67
     Crush
    -0.63
     Rescue
    -0.62
    english
    -0.61
    rency
    -0.60
     Destination
    -0.60
     Sett
    -0.60
    fac
    -0.60
    POSITIVE LOGITS
     belong
    0.83
    isphere
    0.77
    folk
    0.75
     constitute
    0.71
    ensitive
    0.70
    cient
    0.68
     overl
    0.68
    isations
    0.68
    hips
    0.66
    folio
    0.65
    Activations Density 0.479%

    No Known Activations