© 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 EvalsExportsAPI 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. 801
    Prev
    Next
    INDEX
    Explanations

    binary patterns or sequences that repeat at different scales

    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
    640
    +0.47
    2.0%
    160
    +0.14
    0.6%
    167
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.47
    0.39
    160
    +0.14
    0.23
    6
    +0.13
    0.21
    Negative Logits
    irtual
    -0.85
     Mobil
    -0.84
     withdrawal
    -0.80
    ertodd
    -0.79
     scrap
    -0.79
     totaled
    -0.77
     helicop
    -0.77
     amnesty
    -0.76
     respects
    -0.74
     terminating
    -0.73
    POSITIVE LOGITS
    Reviewer
    1.15
    Contents
    1.13
    Rated
    1.10
    Posted
    1.09
    ccording
    1.09
    Edit
    1.07
    PRESS
    1.05
    ³³³³³³³³³³³³³³³³
    1.05
    Synopsis
    1.02
    Disclaimer
    1.01
    Activations Density 2.990%

    No Known Activations