© 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 E2E Recon
    9. 6-RES_SCEFR-AJT
    11. 3476
    Prev
    Next
    INDEX
    Explanations

    legal and judicial terms or phrases

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scefr-ajt/6-res_scefr-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
    140
    +0.13
    0.5%
    197
    +0.12
    0.5%
    750
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    140
    +0.13
    0.04
    594
    +0.12
    0.02
    6
    +0.11
    0.03
    Negative Logits
    icrobial
    -0.92
    zsche
    -0.87
    zbollah
    -0.84
    ongs
    -0.84
    brates
    -0.83
    ramids
    -0.81
    glers
    -0.80
    agues
    -0.78
    brate
    -0.78
    iste
    -0.77
    POSITIVE LOGITS
     358
    0.83
     1280
    0.81
    mingham
    0.80
     361
    0.76
     359
    0.76
     458
    0.73
     Cruiser
    0.70
    JR
    0.70
    Attach
    0.70
     POW
    0.69
    Activations Density 0.204%

    No Known Activations