© 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. 33974
    Prev
    Next
    INDEX
    Explanations

    references to revisiting or updating previous information

    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
    491
    +0.02
    0.5%
    644
    +0.02
    0.4%
    258
    +0.02
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.02
    0.25
    362
    +0.02
    0.14
    393
    +0.02
    0.34
    Negative Logits
     Semin
    -0.14
     theoret
    -0.13
     Supporters
    -0.13
     Shia
    -0.12
    aden
    -0.12
     indisp
    -0.12
    ofi
    -0.12
     seller
    -0.12
     Filipino
    -0.11
     behav
    -0.11
    POSITIVE LOGITS
     revisit
    0.17
    ited
    0.17
    ITED
    0.17
    itals
    0.16
    ilings
    0.16
    ibly
    0.16
    rences
    0.15
    ibilities
    0.15
    ighth
    0.15
    itor
    0.14
    Activations Density 20.168%

    No Known Activations