© 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. 19954
    Prev
    Next
    INDEX
    Explanations

    phrases related to evaluating or considering worth or value

    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
    435
    +0.13
    0.6%
    377
    +0.11
    0.5%
    481
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    435
    +0.13
    0.03
    665
    +0.11
    0.04
    726
    +0.10
    0.03
    Negative Logits
    reated
    -0.68
    idal
    -0.61
     denying
    -0.59
     occupancy
    -0.59
     lett
    -0.57
    orate
    -0.56
    azo
    -0.55
    aga
    -0.54
    Dim
    -0.54
     hostile
    -0.54
    POSITIVE LOGITS
     ASAP
    0.95
     someday
    0.84
     nonetheless
    0.76
     lest
    0.75
     anytime
    0.74
     anew
    0.71
     reconsider
    0.71
     revisit
    0.70
     GOODMAN
    0.69
    !--
    0.68
    Activations Density 0.150%

    No Known Activations