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. 45458
    Prev
    Next
    INDEX
    Explanations

    titles or names of objects or characters with specific color descriptors, such as "Blue Steel" or "Gray Fortress"

    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
    659
    +0.09
    0.8%
    314
    +0.07
    0.6%
    447
    +0.07
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.09
    0.19
    314
    +0.07
    0.00
    213
    +0.07
    0.11
    Negative Logits
     Rebels
    -0.42
     Nirvana
    -0.42
     following
    -0.41
     pret
    -0.39
     TSA
    -0.37
     contr
    -0.36
     Democr
    -0.36
     interviewing
    -0.35
     NDP
    -0.35
     Ebola
    -0.34
    POSITIVE LOGITS
    ocking
    0.65
    amed
    0.61
    amps
    0.61
    ipp
    0.60
    aff
    0.60
    illed
    0.59
    cknowled
    0.57
    aming
    0.57
    aches
    0.57
    ona
    0.56
    Activations Density 8.414%

    No Known Activations