© 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 · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 2800
    Prev
    Next
    INDEX
    Explanations

    detailed descriptions of violent incidents and accidents

    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
    393
    +0.19
    0.9%
    481
    +0.15
    0.7%
    362
    +0.14
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    385
    +0.19
    0.09
    362
    +0.15
    0.08
    608
    +0.14
    0.08
    Negative Logits
    ceive
    -0.79
    Boost
    -0.79
    hei
    -0.78
    omics
    -0.78
    peak
    -0.76
    hops
    -0.75
    2020
    -0.74
    TPP
    -0.74
    leeve
    -0.73
    aber
    -0.72
    POSITIVE LOGITS
     whereabouts
    1.12
     youngest
    1.10
     latter
    1.08
    resa
    1.07
     incident
    1.07
     remainder
    1.06
    odore
    1.06
     Associated
    1.06
     coroner
    1.04
     earliest
    1.02
    Activations Density 0.353%

    No Known Activations