© 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 768
    9. 8-RES_FS768-JB
    11. 650
    Prev
    Next
    INDEX
    Explanations

    information related to locations, events, and organizations

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs768-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    768
    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
    762
    +0.10
    0.5%
    575
    +0.10
    0.5%
    176
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.10
    0.34
    575
    +0.10
    0.30
    547
    +0.10
    0.16
    Negative Logits
    yss
    -0.58
    ãĥ³ãĤ¸
    -0.58
    aeper
    -0.57
    ameless
    -0.54
     pione
    -0.53
    Pitt
    -0.49
    okin
    -0.47
    Els
    -0.47
    adapt
    -0.46
    nai
    -0.46
    POSITIVE LOGITS
    .
    1.39
    *.
    1.28
    .[
    1.27
     respectively
    1.24
    .(
    1.16
    .</
    1.09
    ;
    1.08
    ãĢĤ
    1.08
    .''.
    1.07
    .;
    1.06
    Activations Density 9.755%

    No Known Activations