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

    Neuronpedia

    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. 19709
    Prev
    Next
    INDEX
    Explanations

    mentions of the letter 'h'

    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
    119
    +0.12
    0.5%
    220
    +0.10
    0.5%
    240
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    220
    +0.12
    0.02
    429
    +0.10
    0.02
    119
    +0.10
    0.02
    Negative Logits
    ãĤ´ãĥ³
    -0.79
    terday
    -0.71
    lihood
    -0.68
     mine
    -0.66
    ãĥ¯
    -0.65
     Beware
    -0.65
    Leaks
    -0.63
     flush
    -0.61
     succeeding
    -0.59
    Ô
    -0.59
    POSITIVE LOGITS
    oused
    1.28
    ulk
    1.19
    anging
    1.17
    anky
    1.16
    ospital
    1.15
    ousing
    1.13
    ollywood
    1.12
    olly
    1.12
    uddled
    1.11
    acking
    1.11
    Activations Density 0.015%

    No Known Activations