© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 3928
    Prev
    Next
    INDEX
    Explanations

    words related to social benefits, particularly in contexts like government programs and policy

    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
    Neuron Alignment
    Index
    Value
    % of L₁
    288
    +0.17
    0.8%
    420
    +0.12
    0.6%
    97
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    444
    +0.17
    0.02
    712
    +0.12
    0.02
    285
    +0.11
    0.02
    Negative Logits
     Rim
    -0.75
    ãĥ£
    -0.70
     IPM
    -0.67
    ãĥĥãĥī
    -0.67
     Puzzles
    -0.66
    canon
    -0.65
    swick
    -0.64
     Penal
    -0.63
    bat
    -0.63
     Reloaded
    -0.62
    POSITIVE LOGITS
     conferred
    1.17
    benefit
    0.98
     accrued
    0.97
     afforded
    0.97
    ocratic
    0.94
     claimants
    0.93
     outwe
    0.88
    itism
    0.85
    ocrats
    0.85
     bestowed
    0.85
    Activations Density 0.027%

    No Known Activations