© 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. 15881
    Prev
    Next
    INDEX
    Explanations

    phrases related to police work and law enforcement activities

    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
    590
    +0.10
    0.4%
    509
    +0.10
    0.4%
    57
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    509
    +0.10
    0.03
    617
    +0.10
    0.03
    538
    +0.09
    0.03
    Negative Logits
    yip
    -0.95
    Downloadha
    -0.82
    igible
    -0.81
    xual
    -0.81
    opian
    -0.78
    arget
    -0.75
    å£
    -0.70
    bp
    -0.67
    zsche
    -0.66
    idious
    -0.66
    POSITIVE LOGITS
     Department
    1.35
     Dept
    1.30
     Officers
    1.24
     Officer
    1.14
     Commissioner
    1.13
     Constable
    1.08
     Chief
    1.06
     Bureau
    1.02
     department
    1.00
     departments
    0.98
    Activations Density 0.049%

    No Known Activations