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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 32660
    Prev
    Next
    INDEX
    Explanations

    quotations or direct speech

    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
    +0.16
    0.9%
    314
    +0.11
    0.6%
    393
    +0.09
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.16
    0.19
    160
    +0.11
    0.14
    314
    +0.09
    -0.03
    Negative Logits
     enlisted
    -0.57
     conj
    -0.55
     others
    -0.55
     Vaugh
    -0.54
    ismo
    -0.53
    ij士
    -0.52
    aneers
    -0.51
     folded
    -0.51
     reprint
    -0.50
     enraged
    -0.49
    POSITIVE LOGITS
    yss
    0.81
    Cause
    0.75
    -'
    0.75
    Very
    0.73
    Too
    0.71
    affer
    0.70
    race
    0.70
    Special
    0.69
    SN
    0.68
    Political
    0.66
    Activations Density 4.832%

    No Known Activations