© 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. 11031
    Prev
    Next
    INDEX
    Explanations

    phrases related to advocating for specific beliefs or causes

    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
    393
    +0.32
    1.1%
    385
    +0.23
    0.8%
    718
    +0.22
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    49
    +0.32
    0.10
    718
    +0.23
    0.10
    608
    +0.22
    0.08
    Negative Logits
     looms
    -0.97
    mails
    -0.94
    listed
    -0.90
     Bounty
    -0.85
     Bengal
    -0.84
    chains
    -0.83
    ihad
    -0.82
     warranties
    -0.82
    cuts
    -0.81
     versus
    -0.81
    POSITIVE LOGITS
     minimize
    1.42
    asty
    1.40
     provide
    1.40
     reproduce
    1.39
     conserve
    1.39
     dismantle
    1.38
     educate
    1.37
     counteract
    1.36
     avoid
    1.34
     maximize
    1.34
    Activations Density 0.791%

    No Known Activations