© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 14702
    Prev
    Next
    INDEX
    Explanations

    phrases related to routine or regular activities

    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
    36
    +0.11
    0.5%
    27
    +0.11
    0.5%
    572
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    36
    +0.11
    0.05
    288
    +0.11
    0.06
    572
    +0.11
    0.04
    Negative Logits
    akings
    -0.99
    blems
    -0.93
    assies
    -0.91
    hops
    -0.86
    andals
    -0.84
    casts
    -0.83
    missing
    -0.82
    urches
    -0.82
    ravings
    -0.82
    features
    -0.81
    POSITIVE LOGITS
     standpoint
    1.47
     perspective
    1.23
     basis
    1.19
     viewpoint
    0.98
     mindset
    0.95
     vantage
    0.93
     timetable
    0.92
     manner
    0.92
     timeframe
    0.91
     angle
    0.89
    Activations Density 0.310%

    No Known Activations