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. 34295
    Prev
    Next
    INDEX
    Explanations

    mentions of sports fans and their reactions

    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%
    63
    +0.11
    0.5%
    86
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    522
    +0.26
    0.10
    353
    +0.11
    0.08
    241
    +0.11
    0.08
    Negative Logits
    lished
    -0.72
    nom
    -0.71
    ateral
    -0.68
    ten
    -0.65
    ered
    -0.65
    olon
    -0.64
    Chain
    -0.64
    redo
    -0.64
    otomy
    -0.62
    toe
    -0.62
    POSITIVE LOGITS
     boo
    1.00
     cheering
    0.96
     chanting
    0.90
     chanted
    0.88
    haw
    0.87
     stadiums
    0.86
     Atmosp
    0.86
     cheered
    0.85
    verson
    0.85
    atics
    0.82
    Activations Density 8.750%

    No Known Activations