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. 33701
    Prev
    Next
    INDEX
    Explanations

    financial figures and economic terms

    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
    407
    +0.14
    0.6%
    165
    +0.13
    0.5%
    599
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.14
    0.24
    407
    +0.13
    0.09
    193
    +0.11
    0.10
    Negative Logits
    odox
    -0.84
    heter
    -0.79
     throats
    -0.70
    lev
    -0.69
     dissenting
    -0.66
    fing
    -0.64
     rhythms
    -0.64
    arenthood
    -0.63
    eper
    -0.63
    ensen
    -0.63
    POSITIVE LOGITS
     USD
    1.04
     annually
    0.96
    -$
    0.95
     (£
    0.93
     dollars
    0.89
     AUD
    0.88
    =$
    0.88
    /$
    0.88
    USD
    0.87
     Bonds
    0.87
    Activations Density 11.249%

    No Known Activations