© 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. 8-RES-JB
    11. 11927
    Prev
    Next
    INDEX
    Explanations

    references to names, particularly the name "Earl"

    oai_token-act-pair · gpt-3.5-turbo

    references to individuals named Earl or Darrell

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Atl
    -0.73
    apest
    -0.70
    Pak
    -0.69
    oglu
    -0.68
    okers
    -0.66
    ual
    -0.65
    CW
    -0.65
    ruction
    -0.65
    ocial
    -0.65
    lus
    -0.63
    POSITIVE LOGITS
     lapt
    0.85
    è£ħ
    0.77
    ĪĴ
    0.76
    istry
    0.74
     chairs
    0.73
    ttes
    0.72
    anguage
    0.69
     surv
    0.68
    sworth
    0.68
    izabeth
    0.66
    Activations Density 0.100%

    No Known Activations