© 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 49152
    9. 8-RES_FS49152-JB
    11. 25796
    Prev
    Next
    INDEX
    Explanations

    references to legal or regulatory concepts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    128
    +0.13
    0.6%
    742
    +0.11
    0.5%
    32
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    128
    +0.13
    0.01
    32
    +0.11
    0.01
    556
    +0.09
    0.01
    Negative Logits
     Fey
    -0.76
    HAEL
    -0.75
    olicy
    -0.67
     Moonlight
    -0.63
    plan
    -0.62
     Measure
    -0.62
    BOX
    -0.61
    chnology
    -0.61
     Marketplace
    -0.60
     Ole
    -0.59
    POSITIVE LOGITS
    igion
    1.50
    iever
    1.44
    ief
    1.28
    ieved
    1.26
    iable
    1.22
    ieving
    1.20
    inqu
    1.20
    iance
    1.20
    ievers
    1.18
    atively
    1.17
    Activations Density 0.005%

    No Known Activations