© 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. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    5. GPT2-Small
    7. Residual Similar CE Local
    9. 6-RES_SCL-AJT
    11. 33047
    Prev
    Next
    INDEX
    Explanations

    references to specific individuals and political events related to political campaigns and investigations

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-ajt
    Prompts (Dashboard)
    12,288 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    46,080
    Data Type
    torch.float32
    Hook Point
    blocks.6.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    apollo-research/Skylion007-openwebtext-tokenizer-gpt2
    Hook Point Layer
    6
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    447
    +0.17
    0.8%
    288
    +0.16
    0.8%
    224
    +0.14
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    224
    +0.17
    0.10
    442
    +0.16
    0.08
    336
    +0.14
    0.05
    Negative Logits
    pine
    -0.80
    hua
    -0.77
    eem
    -0.75
    lasses
    -0.74
    ollow
    -0.73
    awks
    -0.72
    ĪĴ
    -0.71
    erd
    -0.70
    erva
    -0.69
    ppo
    -0.66
    POSITIVE LOGITS
    afort
    0.83
     Manafort
    0.71
     Yanukovych
    0.69
    oulos
    0.63
    ufact
    0.62
     coils
    0.59
     buckle
    0.57
    iola
    0.57
    opoulos
    0.56
     annex
    0.56
    Activations Density 7.339%

    No Known Activations