© 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 49152
    9. 8-RES_FS49152-JB
    11. 41728
    Prev
    Next
    INDEX
    Explanations

    terms related to accountability and scrutiny in social and legal contexts

    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
    456
    +0.13
    0.6%
    248
    +0.12
    0.5%
    59
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    59
    +0.13
    0.03
    761
    +0.12
    0.03
    611
    +0.10
    0.03
    Negative Logits
    ãĤ´ãĥ³
    -0.61
    zai
    -0.57
    noon
    -0.56
    _>
    -0.55
     flap
    -0.54
    utt
    -0.54
    pmwiki
    -0.53
    bringer
    -0.53
     relent
    -0.52
     opp
    -0.52
    POSITIVE LOGITS
     us
    1.11
     oneself
    1.06
     him
    0.99
     ourselves
    0.94
     yourselves
    0.93
     individuals
    0.92
     them
    0.91
     themselves
    0.91
    him
    0.90
     consumers
    0.85
    Activations Density 0.129%

    No Known Activations