© 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 12288
    9. 8-RES_FS12288-JB
    11. 742
    Prev
    Next
    INDEX
    Explanations

    places or situations related to vulnerability and basic needs, such as housing, access to services, and support for those in difficult circumstances

    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
    314
    +0.13
    0.6%
    264
    +0.11
    0.5%
    504
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    504
    +0.13
    0.04
    264
    +0.11
    0.02
    737
    +0.11
    0.03
    Negative Logits
     Cly
    -0.70
    requires
    -0.63
    UTH
    -0.62
     Causes
    -0.60
     RELE
    -0.57
     Truman
    -0.56
     equation
    -0.56
    uay
    -0.55
     Beats
    -0.55
     Parallel
    -0.55
    POSITIVE LOGITS
     refuge
    0.86
     dwell
    0.78
     residence
    0.78
    Discussion
    0.74
    quartered
    0.73
    aign
    0.71
    ocate
    0.70
    rawler
    0.69
     worship
    0.69
     desks
    0.68
    Activations Density 0.298%

    No Known Activations