© 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. 8902
    Prev
    Next
    INDEX
    Explanations

    nitrous-related terms and words containing 'rous'

    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
    Neuron Alignment
    Index
    Value
    % of L₁
    527
    +0.06
    0.5%
    316
    +0.06
    0.4%
    93
    +0.05
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.06
    0.06
    288
    +0.06
    0.07
    404
    +0.05
    0.04
    Negative Logits
    opsy
    -0.54
    oby
    -0.43
    stadt
    -0.42
    imates
    -0.41
    olitan
    -0.41
    alach
    -0.40
    aghan
    -0.40
    erno
    -0.40
    atana
    -0.40
    framework
    -0.39
    POSITIVE LOGITS
    atile
    0.59
     oxide
    0.58
    atility
    0.49
     chloride
    0.45
    theless
    0.45
    hes
    0.45
     sights
    0.43
    xual
    0.42
    uled
    0.41
     tremend
    0.41
    Activations Density 0.584%

    No Known Activations