© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 17341
    Prev
    Next
    INDEX
    Explanations

    mentions of personal stories and experiences, especially related to personal beliefs and past events

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    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

    Neuron Alignment
    Index
    Value
    % of L₁
    112
    +0.09
    0.4%
    297
    +0.09
    0.4%
    271
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    271
    +0.09
    0.04
    449
    +0.09
    0.04
    501
    +0.08
    0.04
    Negative Logits
    PDF
    -0.69
     ®
    -0.66
    âĦ¢:
    -0.64
    AU
    -0.63
     Droid
    -0.60
    éĹĺ
    -0.60
    iren
    -0.60
    âī
    -0.59
    aily
    -0.59
    caps
    -0.58
    POSITIVE LOGITS
     secondly
    1.26
     luckily
    1.06
    romeda
    1.05
     then
    1.05
     furthermore
    1.04
     yeah
    1.03
     everybody
    0.99
     I
    0.99
     THEN
    0.95
     fortunately
    0.91
    Activations Density 0.145%

    No Known Activations