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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 37319
    Prev
    Next
    INDEX
    Explanations

    negative or unhelpful aspects related to software functionality

    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
    102
    +0.12
    0.5%
    3
    +0.12
    0.5%
    640
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    152
    +0.12
    0.05
    102
    +0.12
    0.05
    314
    +0.12
    0.01
    Negative Logits
     Gallup
    -0.84
     Polit
    -0.84
     Amnesty
    -0.82
     Phelps
    -0.80
     homicide
    -0.78
     Parish
    -0.77
     Klu
    -0.76
     Ox
    -0.76
     Municipal
    -0.76
     Parenthood
    -0.75
    POSITIVE LOGITS
    compatible
    1.28
    mode
    1.21
    interface
    1.20
    package
    1.15
    level
    1.12
    cache
    1.12
    layer
    1.12
    layout
    1.11
    usable
    1.11
    based
    1.10
    Activations Density 0.068%

    No Known Activations