© 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 1536
    9. 8-RES_FS1536-JB
    11. 283
    Prev
    Next
    INDEX
    Explanations

    sentences with expressions of possibility and desire

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    451
    +0.11
    0.5%
    402
    +0.11
    0.5%
    602
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    763
    +0.11
    0.07
    566
    +0.11
    0.08
    718
    +0.10
    0.09
    Negative Logits
    odic
    -0.62
     groove
    -0.61
     itch
    -0.60
    nexus
    -0.59
     disinfect
    -0.57
     dramas
    -0.57
    umin
    -0.57
    indal
    -0.56
     fascination
    -0.56
    ench
    -0.55
    POSITIVE LOGITS
    would
    0.97
    Had
    0.97
     wouldn
    0.94
     Wouldn
    0.87
     Would
    0.84
     would
    0.83
     sooner
    0.78
     prevented
    0.76
    ivably
    0.76
     Had
    0.75
    Activations Density 2.762%

    No Known Activations