© 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. 9469
    Prev
    Next
    INDEX
    Explanations

    personal pronouns and possessive adjectives associated with confrontational situations

    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
    762
    +0.14
    0.6%
    241
    +0.12
    0.5%
    481
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    762
    +0.14
    0.07
    351
    +0.12
    0.04
    241
    +0.10
    0.04
    Negative Logits
    Balt
    -0.78
    icion
    -0.74
    Americ
    -0.72
    DCS
    -0.71
    INFO
    -0.65
    idth
    -0.62
    ulton
    -0.62
    Wilson
    -0.61
    irection
    -0.60
     advisory
    -0.60
    POSITIVE LOGITS
    selves
    1.31
    self
    1.12
     senseless
    0.97
     selves
    0.93
     uncond
    0.90
     sacrific
    0.88
     anyway
    0.85
     dearly
    0.85
     anymore
    0.82
     verbally
    0.80
    Activations Density 0.162%

    No Known Activations