© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 12-RES-JB
    11. 2651
    Prev
    Next
    INDEX
    Explanations

    phrases related to expressing gratitude or making requests in a formal manner

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot

    archaic or poetic expressions of longing and devotion

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 12-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post
    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_post
    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
    Actually
    -0.67
    activ
    -0.64
    Prime
    -0.63
    ertodd
    -0.62
    Basically
    -0.62
     IPM
    -0.60
    Karl
    -0.60
     basically
    -0.59
    usional
    -0.59
    Newsletter
    -0.59
    POSITIVE LOGITS
     thy
    0.98
     faire
    0.90
     thou
    0.88
     hath
    0.88
     oft
    0.86
     thee
    0.85
     redes
    0.79
     Thy
    0.78
     renown
    0.77
     foe
    0.77
    Activations Density 0.146%

    No Known Activations