© 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. 6274
    Prev
    Next
    INDEX
    Explanations

    words related to characteristics or qualities, particularly negative traits or processes

    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
    287
    +0.10
    0.5%
    543
    +0.10
    0.4%
    233
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    331
    +0.10
    0.05
    616
    +0.10
    0.04
    118
    +0.09
    0.04
    Negative Logits
     designation
    -0.75
     waivers
    -0.74
     Consent
    -0.70
     Generator
    -0.66
     Dragons
    -0.65
     Wolves
    -0.63
     Writers
    -0.63
     reconciliation
    -0.62
     Prevention
    -0.62
     lawsuit
    -0.62
    POSITIVE LOGITS
    istent
    1.15
    erous
    1.10
    ensical
    1.08
    icky
    1.07
    iable
    1.04
    istic
    1.03
    ceptive
    1.03
    hesive
    1.02
    icative
    1.02
    isable
    1.01
    Activations Density 0.273%

    No Known Activations