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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 9248
    Prev
    Next
    INDEX
    Explanations

    numbered lists or sequences

    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
    697
    +0.10
    0.5%
    415
    +0.10
    0.5%
    430
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    415
    +0.10
    0.03
    384
    +0.10
    0.02
    241
    +0.10
    0.02
    Negative Logits
    chy
    -0.77
    etimes
    -0.73
    pers
    -0.67
     minded
    -0.63
    ibilities
    -0.63
    Ward
    -0.63
    ement
    -0.62
    aned
    -0.62
    ITNESS
    -0.61
    ivities
    -0.61
    POSITIVE LOGITS
     onwards
    0.85
     onward
    0.84
     inher
    0.68
     Presents
    0.64
     Forth
    0.60
     Goth
    0.60
     srfAttach
    0.59
     cru
    0.58
    agall
    0.57
    bottom
    0.57
    Activations Density 0.062%

    No Known Activations