© 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. 14208
    Prev
    Next
    INDEX
    Explanations

    proper nouns such as names of people or organizations

    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
    238
    +0.23
    0.7%
    288
    +0.21
    0.6%
    476
    +0.15
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    182
    +0.23
    0.08
    293
    +0.21
    0.08
    316
    +0.15
    0.07
    Negative Logits
     Hitman
    -1.13
     Tables
    -1.11
     Locked
    -1.10
     Picture
    -1.08
     Reborn
    -1.06
    ãĢIJ
    -1.05
     Continued
    -1.02
     Wildcats
    -1.01
     Citation
    -1.00
     Sins
    -0.99
    POSITIVE LOGITS
    monary
    1.78
    iple
    1.51
    enium
    1.43
    ums
    1.42
    terior
    1.38
    pees
    1.38
    gement
    1.35
    esc
    1.33
    umin
    1.33
    gence
    1.30
    Activations Density 2.124%

    No Known Activations