© 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 3072
    9. 8-RES_FS3072-JB
    11. 25
    Prev
    Next
    INDEX
    Explanations

    references to videos

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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
    571
    +0.10
    0.4%
    516
    +0.10
    0.4%
    540
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    690
    +0.10
    0.05
    516
    +0.10
    0.04
    540
    +0.09
    0.05
    Negative Logits
    Ĭ±
    -0.83
    姫
    -0.76
    owder
    -0.74
    uckle
    -0.71
    cffff
    -0.70
    erella
    -0.68
    sembly
    -0.68
    vier
    -0.68
    izabeth
    -0.66
    vironment
    -0.65
    POSITIVE LOGITS
     clips
    1.22
     clip
    1.22
     footage
    1.19
    clip
    1.07
     filmed
    1.00
     recordings
    0.96
     uploaded
    0.94
     recorder
    0.90
     recording
    0.90
     cameras
    0.89
    Activations Density 0.402%

    No Known Activations