© 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. 18038
    Prev
    Next
    INDEX
    Explanations

    mentions of specific companies and their 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
    534
    +0.10
    0.5%
    640
    +0.09
    0.4%
    246
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    632
    +0.10
    0.05
    534
    +0.09
    0.04
    626
    +0.08
    0.04
    Negative Logits
     pus
    -0.85
     pim
    -0.62
     advis
    -0.62
     steering
    -0.62
     resusc
    -0.60
     drill
    -0.60
     heroism
    -0.59
     acceler
    -0.58
     organizational
    -0.58
     parasites
    -0.57
    POSITIVE LOGITS
    $.
    1.16
    Va
    0.99
    S
    0.97
     Territories
    0.81
    England
    0.80
     States
    0.77
    western
    0.77
    ¢
    0.77
     Nations
    0.76
     STATES
    0.75
    Activations Density 0.040%

    No Known Activations