© 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 Local
    9. 6-RES_SCL-AJT
    11. 40631
    Prev
    Next
    INDEX
    Explanations

    pieces of code

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-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
    447
    +0.11
    0.8%
    393
    +0.07
    0.5%
    176
    +0.06
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    255
    +0.11
    0.14
    570
    +0.07
    -0.01
    698
    +0.06
    0.10
    Negative Logits
    laus
    -0.44
    uits
    -0.43
    artments
    -0.43
     supervision
    -0.42
    itarian
    -0.41
     restraints
    -0.41
     jurisdictions
    -0.40
     leash
    -0.40
     Anton
    -0.39
    izon
    -0.39
    POSITIVE LOGITS
     output
    0.53
    DragonMagazine
    0.52
     outputs
    0.51
     indistinguishable
    0.50
     Output
    0.48
     payload
    0.45
     milliseconds
    0.45
    output
    0.45
    bytes
    0.43
     gib
    0.43
    Activations Density 17.895%

    No Known Activations