© 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 E2E Recon
    9. 6-RES_SCEFR-AJT
    11. 14140
    Prev
    Next
    INDEX
    Explanations

    phrases related to tax policies

    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
    393
    +0.22
    0.7%
    311
    +0.15
    0.5%
    526
    +0.14
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    471
    +0.22
    0.09
    90
    +0.15
    0.08
    654
    +0.14
    0.08
    Negative Logits
    SourceFile
    -1.25
    soDeliveryDate
    -1.10
    olf
    -1.08
    ¯¯
    -1.08
    leground
    -1.07
    Champ
    -1.07
    Allah
    -1.05
    #$
    -1.00
    ikan
    -0.97
    AppData
    -0.97
    POSITIVE LOGITS
     behalf
    1.91
    erous
    1.91
    etime
    1.35
     Sundays
    1.26
     arrival
    1.24
     Saturdays
    1.22
     Mondays
    1.19
     Fridays
    1.15
    yx
    1.13
     sale
    1.11
    Activations Density 1.428%

    No Known Activations