© 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 49152
    9. 8-RES_FS49152-JB
    11. 45717
    Prev
    Next
    INDEX
    Explanations

    references to the concept of "magic" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    288
    +0.19
    0.9%
    642
    +0.11
    0.5%
    48
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.19
    0.01
    32
    +0.11
    0.01
    48
    +0.10
    0.01
    Negative Logits
    akeru
    -0.78
    arers
    -0.76
    aeper
    -0.74
    bis
    -0.74
     Fas
    -0.71
    rontal
    -0.70
    fam
    -0.70
    Filename
    -0.70
    upon
    -0.70
    alis
    -0.67
    POSITIVE LOGITS
     wand
    0.98
     tricks
    0.97
     realism
    0.90
     lantern
    0.87
     mushrooms
    0.84
     carpet
    0.84
    ãĥĥãĤ¯
    0.83
     potion
    0.82
     beans
    0.81
     trick
    0.79
    Activations Density 0.008%

    No Known Activations