© 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. 9093
    Prev
    Next
    INDEX
    Explanations

    keywords related to investigative actions, such as searching and digging

    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
    569
    +0.12
    0.5%
    140
    +0.09
    0.4%
    383
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    569
    +0.12
    0.02
    113
    +0.09
    0.02
    140
    +0.09
    0.02
    Negative Logits
    BuyableInstoreAndOnline
    -0.69
     Leilan
    -0.67
    âĹ¼
    -0.64
    cia
    -0.63
    âĢ¢âĢ¢
    -0.60
    Number
    -0.59
     constitu
    -0.58
    --------------------------------
    -0.58
    ---------
    -0.58
    channelAvailability
    -0.58
    POSITIVE LOGITS
     trenches
    1.15
     dug
    1.04
    itial
    0.99
     digging
    0.99
     graves
    0.98
    ression
    0.94
     dig
    0.93
     digs
    0.93
     deeper
    0.92
     holes
    0.92
    Activations Density 0.019%

    No Known Activations