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 E2E Recon
    5. 6-RES_SCEFR-AJT
    6. 493
    Prev
    Next
    INDEX
    Explanations

    numeric values or special terms presented in a specific format (e.g., wavelengths, atomic measurements, etc.)

    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
    0
    +0.14
    0.5%
    151
    +0.14
    0.5%
    101
    +0.14
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    151
    +0.14
    0.14
    422
    +0.14
    0.09
    106
    +0.14
    0.10
    Negative Logits
    ĺħ
    -1.03
     Majority
    -0.98
    retty
    -0.95
     Lessons
    -0.88
     Chef
    -0.87
    eto
    -0.87
     Hardware
    -0.86
     Techniques
    -0.85
    uese
    -0.85
     Journals
    -0.84
    POSITIVE LOGITS
     spaced
    1.07
     combust
    1.05
     priced
    1.00
     differentiated
    0.98
     populated
    0.96
    protected
    0.96
    encrypted
    0.93
    ppy
    0.93
     localized
    0.92
    pless
    0.92
    Activations Density 2.083%

    No Known Activations