© 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. 15880
    Prev
    Next
    INDEX
    Explanations

    phrases related to giving opinions or reporting statements from officials

    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
    283
    +0.09
    0.4%
    524
    +0.09
    0.4%
    655
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    283
    +0.09
    0.03
    431
    +0.09
    0.03
    86
    +0.09
    0.03
    Negative Logits
     Ammo
    -0.75
    orio
    -0.61
    ptions
    -0.61
     Vine
    -0.58
    earances
    -0.56
     Revival
    -0.56
     Conversation
    -0.55
     Engineers
    -0.55
     Arlington
    -0.55
     Predators
    -0.55
    POSITIVE LOGITS
     resembles
    0.92
     constitutes
    0.84
     ought
    0.82
     violates
    0.82
     resembled
    0.80
     represents
    0.80
     amounted
    0.78
     originate
    0.77
     signifies
    0.77
     deems
    0.77
    Activations Density 0.129%

    No Known Activations