© 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. 22676
    Prev
    Next
    INDEX
    Explanations

    question-and-answer structures

    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
    373
    +0.18
    0.8%
    507
    +0.13
    0.6%
    292
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    130
    +0.18
    0.01
    656
    +0.13
    0.01
    292
    +0.11
    0.01
    Negative Logits
     blot
    -0.82
     stret
    -0.71
     Juda
    -0.65
     cart
    -0.62
     scales
    -0.61
     Uran
    -0.60
     Sno
    -0.58
    wcs
    -0.56
     Argon
    -0.56
     ascend
    -0.55
    POSITIVE LOGITS
    Answer
    1.06
    orically
    0.78
     Answer
    0.78
    uchin
    0.72
    erate
    0.68
    ause
    0.67
    outheastern
    0.67
    ired
    0.66
    Yes
    0.66
    swers
    0.65
    Activations Density 0.033%

    No Known Activations