Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    3. GPT2-Small
    4. Residual Similar CE Local
    5. 6-RES_SCL-AJT
    6. 45399
    Prev
    Next
    INDEX
    Explanations

    references to specific types of aircraft

    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
    +0.26
    1.1%
    67
    +0.16
    0.6%
    314
    +0.15
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    314
    +0.26
    -0.01
    659
    +0.16
    0.11
    582
    +0.15
    0.07
    Negative Logits
     Gutenberg
    -0.89
     Pharaoh
    -0.70
    wcsstore
    -0.69
     dere
    -0.69
     choke
    -0.67
     Cheong
    -0.65
     Dire
    -0.65
     choir
    -0.64
     Valhalla
    -0.64
     Tycoon
    -0.64
    POSITIVE LOGITS
    odies
    1.29
    rug
    1.27
    asket
    1.27
    amboo
    1.26
    urt
    1.26
    anned
    1.25
    razen
    1.24
    rows
    1.22
    ETA
    1.22
    aked
    1.22
    Activations Density 10.214%

    No Known Activations