© 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. 18911
    Prev
    Next
    INDEX
    Explanations

    phrases related to someone or something returning or being present again

    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
    659
    +0.10
    0.4%
    372
    +0.09
    0.4%
    223
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    372
    +0.10
    0.02
    760
    +0.09
    0.02
    242
    +0.09
    0.02
    Negative Logits
    uary
    -0.64
     Conrad
    -0.63
    imize
    -0.63
     Sorceress
    -0.63
    urations
    -0.61
    "]=>
    -0.61
     Defendants
    -0.59
    AIN
    -0.59
     Consent
    -0.58
     Specific
    -0.58
    POSITIVE LOGITS
    tracking
    1.23
    loaded
    1.01
    ported
    1.00
    ped
    0.97
    packs
    0.96
    ordered
    0.96
    dated
    0.94
    GROUND
    0.94
    stab
    0.93
    track
    0.92
    Activations Density 0.028%

    No Known Activations