© 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
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 24576
    5. 8-RES_FS24576-JB
    6. 24140
    Prev
    Next
    INDEX
    Explanations

    negative sentiments or expressions

    oai_token-act-pair · gpt-4o-miniTriggered 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
    314
    +0.16
    0.7%
    102
    +0.12
    0.6%
    3
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    3
    +0.16
    0.03
    290
    +0.12
    0.02
    102
    +0.11
    0.02
    Negative Logits
     Nig
    -0.71
     Holding
    -0.70
     SHARES
    -0.68
     shards
    -0.67
     DRAG
    -0.67
     Cohn
    -0.67
     streams
    -0.65
     DeV
    -0.64
     merits
    -0.63
     sorely
    -0.63
    POSITIVE LOGITS
    existing
    1.64
    empt
    1.46
    requisite
    1.37
    requisites
    1.34
    release
    1.26
    trial
    1.26
    condition
    1.25
    selected
    1.23
    existent
    1.20
    historic
    1.19
    Activations Density 0.014%

    No Known Activations