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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & 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. 34524
    Prev
    Next
    INDEX
    Explanations

    exact words or phrases that are similar or repeated for emphasis

    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
    651
    +0.18
    0.7%
    447
    +0.18
    0.7%
    151
    +0.14
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    253
    +0.18
    0.13
    536
    +0.18
    0.08
    307
    +0.14
    0.08
    Negative Logits
    ker
    -0.94
    rift
    -0.83
    itiz
    -0.82
    olyn
    -0.80
    kers
    -0.78
    jay
    -0.78
    rug
    -0.74
    roe
    -0.74
    asta
    -0.74
    isson
    -0.73
    POSITIVE LOGITS
    ãĤ¨
    0.89
     opposite
    0.86
     aligned
    0.83
     wrong
    0.83
     matched
    0.75
     suited
    0.75
     positioned
    0.75
    æ©Ł
    0.74
    Els
    0.74
     tuned
    0.71
    Activations Density 8.570%

    No Known Activations