© 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. 2358
    Prev
    Next
    INDEX
    Explanations

    words related to achievements or accomplishments

    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
    549
    +0.17
    0.8%
    314
    +0.12
    0.5%
    16
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    549
    +0.17
    0.02
    701
    +0.12
    0.02
    659
    +0.12
    0.02
    Negative Logits
    ©¶æ
    -0.68
    PDATE
    -0.67
     withd
    -0.62
    drm
    -0.61
     futures
    -0.61
    ences
    -0.60
    fml
    -0.60
     rall
    -0.58
     stocking
    -0.58
    LEASE
    -0.58
    POSITIVE LOGITS
    idon
    1.06
    ieving
    1.04
    annel
    1.03
    akra
    0.99
    auri
    0.96
    iev
    0.93
    alos
    0.93
    ieve
    0.93
    rome
    0.91
    mann
    0.90
    Activations Density 0.029%

    No Known Activations