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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 34092
    Prev
    Next
    INDEX
    Explanations

    words related to sequels, continuations, or follow-ups in various contexts

    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
    +0.25
    1.4%
    274
    +0.12
    0.7%
    549
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    274
    +0.25
    0.11
    549
    +0.12
    0.07
    749
    +0.10
    0.06
    Negative Logits
    ¬¼
    -0.57
    ·
    -0.51
    Ĭ±
    -0.50
    hoff
    -0.49
    uda
    -0.48
     Iranians
    -0.48
    OVER
    -0.48
    urrency
    -0.47
     traders
    -0.47
    WI
    -0.46
    POSITIVE LOGITS
    ngth
    0.70
     sequels
    0.70
     sequel
    0.69
    mort
    0.68
     trilogy
    0.67
     thereto
    0.66
    arium
    0.60
    ogue
    0.60
    arily
    0.58
     installments
    0.58
    Activations Density 6.885%

    No Known Activations