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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 13230
    Prev
    Next
    INDEX
    Explanations

    scientific and technical terms related to surfaces, boundaries, and limits

    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
    +0.22
    1.0%
    288
    +0.15
    0.7%
    537
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    721
    +0.22
    0.08
    268
    +0.15
    0.04
    549
    +0.10
    0.07
    Negative Logits
    ulz
    -0.89
    obe
    -0.75
    orah
    -0.74
    fortune
    -0.73
    ocker
    -0.71
    kus
    -0.70
    enza
    -0.70
    imar
    -0.68
    akia
    -0.68
    aughed
    -0.67
    POSITIVE LOGITS
     boundaries
    1.15
     boundary
    1.03
     markers
    0.85
     separating
    0.85
     crossings
    0.85
     clearance
    0.83
     bounds
    0.82
     margin
    0.76
     Bound
    0.76
     Limits
    0.76
    Activations Density 8.970%

    No Known Activations