© 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. 14685
    Prev
    Next
    INDEX
    Explanations

    mentions of law enforcement activities and procedures

    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
    50
    +0.11
    0.5%
    459
    +0.09
    0.4%
    51
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.11
    0.09
    149
    +0.09
    0.04
    45
    +0.09
    0.03
    Negative Logits
    urther
    -0.80
    grave
    -0.76
    Had
    -0.76
    ajor
    -0.74
    å°Ĩ
    -0.72
    ilateral
    -0.72
    rez
    -0.68
    aclysm
    -0.67
    reed
    -0.67
    expected
    -0.65
    POSITIVE LOGITS
     Occasionally
    1.23
     occasionally
    1.19
    sometimes
    1.13
     sometimes
    1.08
    usually
    1.07
    often
    1.07
     invariably
    1.02
     Sometimes
    1.02
     usually
    1.01
     constantly
    0.98
    Activations Density 0.921%

    No Known Activations