© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 7978
    Prev
    Next
    INDEX
    Explanations

    hexadecimal color codes

    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
    210
    +0.12
    0.6%
    314
    +0.12
    0.5%
    173
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    210
    +0.12
    0.03
    484
    +0.12
    0.03
    445
    +0.10
    0.03
    Negative Logits
     Vie
    -0.78
     Gors
    -0.74
     destro
    -0.70
     Sylv
    -0.70
     cler
    -0.66
     disturbed
    -0.65
     foremost
    -0.65
     Rodrigo
    -0.65
     jog
    -0.64
     Shiva
    -0.64
    POSITIVE LOGITS
    define
    1.29
    DIV
    1.24
    !/
    1.24
    ################################
    1.14
    include
    1.13
    ########
    1.09
    Gamer
    1.07
    ###
    1.06
    DonaldTrump
    0.96
    NAME
    0.96
    Activations Density 0.022%

    No Known Activations