© 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 12288
    9. 8-RES_FS12288-JB
    11. 9437
    Prev
    Next
    INDEX
    Explanations

    phrases or words ending in question marks

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    285
    +0.11
    0.5%
    391
    +0.10
    0.5%
    373
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    285
    +0.11
    0.04
    6
    +0.10
    0.05
    391
    +0.09
    0.03
    Negative Logits
    ality
    -0.69
    apers
    -0.68
     encount
    -0.67
     Nadu
    -0.67
     marsh
    -0.66
    aper
    -0.64
    ulla
    -0.64
     charism
    -0.63
    esian
    -0.63
     conduc
    -0.62
    POSITIVE LOGITS
     Nope
    1.26
    ����
    1.13
    .?
    1.05
     Yeah
    1.00
     Nah
    0.99
    ?:
    0.98
     Huh
    0.98
     Yep
    0.96
     Surely
    0.94
     Why
    0.93
    Activations Density 0.106%

    No Known Activations