© 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 49152
    9. 8-RES_FS49152-JB
    11. 37240
    Prev
    Next
    INDEX
    Explanations

    phrases indicating a sense of loss or absence

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    393
    +0.12
    0.5%
    608
    +0.11
    0.5%
    481
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    669
    +0.12
    0.03
    608
    +0.11
    0.03
    666
    +0.09
    0.02
    Negative Logits
    idelines
    -0.83
    eds
    -0.78
    itures
    -0.77
    hips
    -0.76
    breaks
    -0.75
    Topic
    -0.73
    gemony
    -0.72
    imentary
    -0.69
    hesis
    -0.69
    reements
    -0.69
    POSITIVE LOGITS
     encountering
    0.99
     owning
    0.98
     seeing
    0.96
     having
    0.95
     knowing
    0.93
     watching
    0.89
     discovering
    0.86
     witnessing
    0.85
     childbirth
    0.81
     being
    0.81
    Activations Density 0.083%

    No Known Activations