© 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. 12891
    Prev
    Next
    INDEX
    Explanations

    city names

    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
    373
    +0.21
    1.1%
    314
    +0.11
    0.5%
    519
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    373
    +0.21
    -0.02
    80
    +0.11
    0.04
    545
    +0.11
    0.04
    Negative Logits
    ahime
    -0.67
    nw
    -0.55
     Franch
    -0.54
    kamp
    -0.53
     termin
    -0.48
    eon
    -0.47
    ung
    -0.47
    anges
    -0.46
     priceless
    -0.46
     tresp
    -0.46
    POSITIVE LOGITS
    dden
    0.74
    avis
    0.66
    ody
    0.65
    Marginal
    0.63
    apixel
    0.59
    ciation
    0.59
    ĵĺ
    0.59
    ario
    0.58
    practice
    0.57
    icz
    0.57
    Activations Density 0.352%

    No Known Activations