© 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. 45504
    Prev
    Next
    INDEX
    Explanations

    medical terms and conditions related to health and safety

    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
    373
    +0.16
    0.8%
    218
    +0.14
    0.7%
    124
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.16
    0.03
    659
    +0.14
    0.03
    124
    +0.12
    0.03
    Negative Logits
    etsk
    -0.71
     Reader
    -0.62
     Zap
    -0.61
     Cipher
    -0.60
    ghai
    -0.59
    ãĥ¼ãĥĨ
    -0.59
    baugh
    -0.58
    san
    -0.58
    mass
    -0.57
     wink
    -0.57
    POSITIVE LOGITS
    OUN
    1.17
    OWN
    1.10
    EST
    1.09
    AND
    1.06
    IND
    1.05
     SERVICES
    1.05
    ILL
    1.04
    OUND
    1.04
    AMES
    1.04
    ENT
    1.03
    Activations Density 0.072%

    No Known Activations