© 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 EvalsExportsinterp-engineNEWAPI 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. 22665
    Prev
    Next
    INDEX
    Explanations

    phrases emphasizing collaboration or association

    oai_token-act-pair · gpt-4o-miniTriggered 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
    393
    +0.14
    0.6%
    83
    +0.14
    0.6%
    335
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    83
    +0.14
    0.02
    761
    +0.14
    0.02
    608
    +0.09
    0.03
    Negative Logits
    former
    -0.68
    AU
    -0.65
    BR
    -0.64
    leave
    -0.63
    Congress
    -0.63
    cent
    -0.59
     aloud
    -0.59
    MQ
    -0.59
    usa
    -0.59
    laughter
    -0.58
    POSITIVE LOGITS
     regards
    0.96
     ample
    0.90
    stood
    0.88
    ãĥīãĥ©ãĤ´ãĥ³
    0.88
     sufficient
    0.88
     regard
    0.86
     respect
    0.84
     plenty
    0.83
    ãĤ´ãĥ³
    0.78
    ãĤ¦ãĤ¹
    0.75
    Activations Density 0.077%

    No Known Activations