© 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. 12326
    Prev
    Next
    INDEX
    Explanations

    phrases related to retracting or walking back statements or actions

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to accountability or failure to fulfill promises

    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
     varies
    -0.69
     âĢº
    -0.67
    zees
    -0.67
    teness
    -0.66
     Devi
    -0.64
    ubi
    -0.64
     awa
    -0.64
     Harvest
    -0.64
     GOOD
    -0.63
     HERE
    -0.63
    POSITIVE LOGITS
     filib
    0.94
     denouncing
    0.82
     Nixon
    0.81
     pund
    0.80
     Bannon
    0.79
     neocons
    0.78
     Reagan
    0.78
     impeachment
    0.77
     hastily
    0.76
     lobbying
    0.76
    Activations Density 1.549%

    No Known Activations