© 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. 12647
    Prev
    Next
    INDEX
    Explanations

    phrases related to social and political commentary

    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
    351
    +0.13
    0.6%
    481
    +0.12
    0.6%
    726
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    351
    +0.13
    0.05
    566
    +0.12
    0.06
    526
    +0.10
    -0.02
    Negative Logits
    ãĥ©ãĥ³
    -0.61
     clarify
    -0.59
    hra
    -0.58
     Previous
    -0.57
    cussion
    -0.57
    actionDate
    -0.56
     Initially
    -0.56
     Released
    -0.55
    RFC
    -0.55
     recap
    -0.55
    POSITIVE LOGITS
     goddamn
    0.91
     godd
    0.83
     damn
    0.78
     theirs
    0.76
     crap
    0.73
     darn
    0.72
     indistinguishable
    0.69
     freaking
    0.69
     dude
    0.68
     magically
    0.67
    Activations Density 1.250%

    No Known Activations