© 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 24576
    9. 8-RES_FS24576-JB
    11. 1759
    Prev
    Next
    INDEX
    Explanations

    situations of distress and assault

    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
    373
    +0.14
    0.6%
    545
    +0.11
    0.5%
    730
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    566
    +0.14
    0.07
    718
    +0.11
    0.07
    762
    +0.09
    0.07
    Negative Logits
    senal
    -0.86
    ospons
    -0.85
     Lic
    -0.80
     endors
    -0.79
    tains
    -0.78
     Workshop
    -0.77
    venue
    -0.76
    osponsors
    -0.75
     nominees
    -0.73
    idates
    -0.73
    POSITIVE LOGITS
     numb
    1.26
     sob
    1.24
     panic
    1.21
     imagining
    1.12
     panicked
    1.10
     frantic
    1.09
     dizz
    1.09
     frantically
    1.08
     trembling
    1.04
     crying
    1.04
    Activations Density 0.447%

    No Known Activations