© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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 E2E Recon
    9. 6-RES_SCEFR-AJT
    11. 2312
    Prev
    Next
    INDEX
    Explanations

    phrases related to contrasting ideas or going against established beliefs

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scefr-ajt/6-res_scefr-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
    393
    +0.18
    0.6%
    235
    +0.15
    0.5%
    521
    +0.14
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.18
    0.10
    669
    +0.15
    0.07
    589
    +0.14
    0.06
    Negative Logits
    peat
    -1.07
    bage
    -1.01
    soever
    -0.98
    quel
    -0.98
    wine
    -0.96
    bodied
    -0.95
    cule
    -0.94
    reau
    -0.91
    gebra
    -0.90
    lor
    -0.88
    POSITIVE LOGITS
     tradition
    1.03
     principle
    0.91
    instance
    0.89
     our
    0.89
     orders
    0.87
     stereotype
    0.86
     whom
    0.84
     Iw
    0.83
     directives
    0.83
     precaution
    0.83
    Activations Density 0.997%

    No Known Activations