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

    Neuronpedia

    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. 1805
    Prev
    Next
    INDEX
    Explanations

    references to espionage activities and spy-related terms

    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
    465
    +0.19
    0.7%
    167
    +0.17
    0.6%
    568
    +0.15
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.19
    0.13
    288
    +0.17
    0.14
    465
    +0.15
    0.07
    Negative Logits
    omorphic
    -0.93
    assetsadobe
    -0.92
    disabled
    -0.89
    aucas
    -0.88
     Pradesh
    -0.88
     imbalance
    -0.87
     forc
    -0.85
    omorph
    -0.85
    usable
    -0.84
    xual
    -0.82
    POSITIVE LOGITS
    boys
    1.13
    chu
    1.10
    tail
    1.09
    ry
    1.07
     Claw
    1.06
    iour
    1.06
    sonian
    1.06
     Doodle
    1.05
    glass
    1.04
    meat
    1.04
    Activations Density 2.337%

    No Known Activations