© 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. 8234
    Prev
    Next
    INDEX
    Explanations

    the word "pizza" along with related terms such as "delivery" and "toppings"

    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
    288
    +0.20
    0.6%
    343
    +0.16
    0.5%
    587
    +0.16
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    343
    +0.20
    0.05
    326
    +0.16
    0.05
    288
    +0.16
    0.06
    Negative Logits
    draw
    -1.23
    Thom
    -1.20
    drawn
    -1.14
    say
    -1.11
     Luther
    -1.11
    pring
    -1.06
    productive
    -1.06
    hips
    -1.05
    orne
    -1.05
     Ago
    -1.03
    POSITIVE LOGITS
     crust
    1.50
     dough
    1.47
    ocalypse
    1.40
     oven
    1.35
    agate
    1.31
     pies
    1.28
     pie
    1.24
     Dough
    1.22
    artisan
    1.18
    lyak
    1.15
    Activations Density 0.495%

    No Known Activations