© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 2829
    Prev
    Next
    INDEX
    Explanations

    instances of mistakes or errors

    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
    635
    +0.20
    0.6%
    151
    +0.17
    0.5%
    288
    +0.17
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    151
    +0.20
    0.08
    288
    +0.17
    0.08
    635
    +0.17
    0.05
    Negative Logits
    borough
    -1.10
    rollers
    -1.04
    soon
    -1.03
    iaries
    -1.02
    electric
    -1.01
    iries
    -1.01
    borg
    -1.00
    esc
    -0.99
    ighth
    -0.99
    irie
    -0.99
    POSITIVE LOGITS
     mistakes
    1.35
     mistake
    1.28
     unfocusedRange
    1.14
    ãĤ°
    1.13
    é»Ĵ
    1.08
     omission
    1.06
    ãĥķ
    1.04
     dece
    1.03
     Pelosi
    1.02
    GBT
    1.01
    Activations Density 1.015%

    No Known Activations