© 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 24576
    9. 8-RES_FS24576-JB
    11. 19320
    Prev
    Next
    INDEX
    Explanations

    incidences of medical emergencies and disasters

    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
    480
    +0.12
    0.6%
    481
    +0.12
    0.5%
    696
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    180
    +0.12
    0.03
    478
    +0.12
    0.04
    121
    +0.11
    0.04
    Negative Logits
    ãĥĥãĥī
    -0.78
     sooner
    -0.72
    ¶ħ
    -0.72
     equivalents
    -0.69
    BILITIES
    -0.69
    whatever
    -0.68
     notwithstanding
    -0.68
    govtrack
    -0.68
    ¥µ
    -0.67
    Ħ¢
    -0.65
    POSITIVE LOGITS
     Boko
    0.70
     mural
    0.70
     graffiti
    0.68
     entrance
    0.65
     makeshift
    0.65
     Sochi
    0.65
     Occupy
    0.64
     Mandatory
    0.63
     commem
    0.63
     evict
    0.62
    Activations Density 0.452%

    No Known Activations