© 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. 33960
    Prev
    Next
    INDEX
    Explanations

    comparisons or contrasts between different elements or concepts

    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
    393
    +0.15
    0.7%
    447
    +0.14
    0.6%
    218
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.15
    0.24
    172
    +0.14
    0.10
    385
    +0.11
    0.13
    Negative Logits
    OUR
    -0.69
    âĹ¼
    -0.63
    GES
    -0.62
    eur
    -0.62
    Ire
    -0.61
    Sharp
    -0.60
    ALLY
    -0.57
    ORY
    -0.56
    pez
    -0.56
    chemical
    -0.56
    POSITIVE LOGITS
    pires
    1.12
    pired
    1.10
    bestos
    1.08
    pects
    1.05
    piration
    1.00
    phalt
    1.00
    semble
    1.00
     opposed
    0.99
    piring
    0.96
    semb
    0.96
    Activations Density 10.123%

    No Known Activations