© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    5. GPT2-Small
    7. Residual Similar CE Local
    9. 6-RES_SCL-AJT
    11. 33133
    Prev
    Next
    INDEX
    Explanations

    words related to subscribing, verifying, and selecting options in a text-based form

    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
    211
    +0.08
    0.6%
    604
    +0.07
    0.5%
    302
    +0.07
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.08
    0.17
    36
    +0.07
    0.10
    362
    +0.07
    0.08
    Negative Logits
    geist
    -0.54
    matter
    -0.49
    gew
    -0.49
    imer
    -0.46
    ruck
    -0.45
    farm
    -0.44
    aughs
    -0.44
    illon
    -0.44
     Kov
    -0.43
    Extreme
    -0.43
    POSITIVE LOGITS
     Cancel
    0.52
     login
    0.51
     manually
    0.46
     exit
    0.45
     embed
    0.44
     domains
    0.43
     digits
    0.43
     delete
    0.43
     payment
    0.42
     deletion
    0.42
    Activations Density 4.620%

    No Known Activations