© 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. 10330
    Prev
    Next
    INDEX
    Explanations

    phrases related to importance or urgency

    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
    239
    +0.11
    0.5%
    176
    +0.11
    0.5%
    276
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    737
    +0.11
    0.05
    271
    +0.11
    0.04
    455
    +0.10
    0.04
    Negative Logits
    ĺħ
    -0.74
    CHAT
    -0.66
    essler
    -0.61
    igslist
    -0.61
    ¬¼
    -0.60
    ACY
    -0.58
     teasp
    -0.57
    alsh
    -0.56
    acl
    -0.56
    cot
    -0.56
    POSITIVE LOGITS
     than
    2.00
    than
    1.80
     Than
    1.42
    worldly
    0.82
    erous
    0.80
    iating
    0.78
    iated
    0.75
    ially
    0.70
    ulent
    0.68
    iable
    0.68
    Activations Density 0.169%

    No Known Activations