© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 5838
    Prev
    Next
    INDEX
    Explanations

    verbs related to communication or actions related to sharing information

    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
    715
    +0.11
    0.5%
    43
    +0.10
    0.5%
    70
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    665
    +0.11
    0.06
    70
    +0.10
    0.06
    351
    +0.10
    0.04
    Negative Logits
     Confederation
    -0.60
    she
    -0.59
    tip
    -0.58
    amazon
    -0.58
     Vid
    -0.57
     Flo
    -0.56
     wont
    -0.56
     Wid
    -0.55
     Prin
    -0.54
     Wolver
    -0.54
    POSITIVE LOGITS
     oneself
    1.62
     yourself
    1.14
     ourselves
    1.01
     yourselves
    0.93
     Yourself
    0.79
    entious
    0.79
     somebody
    0.79
    enance
    0.79
    lessly
    0.76
     sacrific
    0.75
    Activations Density 0.285%

    No Known Activations