© 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. 19758
    Prev
    Next
    INDEX
    Explanations

    details related to criminal activities and legal proceedings

    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
    640
    +0.19
    0.9%
    481
    +0.18
    0.8%
    100
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.19
    0.12
    6
    +0.18
    0.10
    160
    +0.12
    0.09
    Negative Logits
     optim
    -0.80
     UX
    -0.79
     sustainability
    -0.79
     ecosystem
    -0.77
     everywhere
    -0.77
     trickle
    -0.77
     migr
    -0.77
     emoji
    -0.76
     usability
    -0.74
     mainline
    -0.72
    POSITIVE LOGITS
    Prosecutors
    1.78
    Prosecut
    1.68
    Court
    1.61
    Judge
    1.53
    Attorney
    1.44
     Prosecutors
    1.40
     defendants
    1.36
     defendant
    1.35
     Defendant
    1.33
    Charges
    1.32
    Activations Density 0.448%

    No Known Activations