© 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. 11-RES-JB
    11. 3537
    Prev
    Next
    INDEX
    Explanations

    phrases expressing hypothetical situations or suggestions

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating propositions or suggestions that begin with "it would be."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    sometimes
    -0.64
    PsyNetMessage
    -0.64
    itus
    -0.61
    anos
    -0.59
    culosis
    -0.59
    Rot
    -0.59
    ickets
    -0.58
    ————
    -0.57
    iop
    -0.57
    ESE
    -0.57
    POSITIVE LOGITS
     preferable
    1.08
     unthinkable
    1.05
     nice
    1.04
     wise
    1.02
     foolish
    1.01
     wiser
    0.93
     impractical
    0.91
     ideal
    0.91
     folly
    0.90
    astrous
    0.89
    Activations Density 0.140%

    No Known Activations