© 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
    2. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    3. GPT2-Small
    4. Residual Similar CE Local
    5. 6-RES_SCL-AJT
    6. 4653
    Prev
    Next
    INDEX
    Explanations

    texts discussing technical details about computer hardware and technology

    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
    •Layer 6 UMAP region: Mostly-local line at bottom - local

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    64
    +0.15
    1.3%
    447
    +0.12
    1.1%
    464
    +0.12
    1.0%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    87
    +0.15
    -0.12
    756
    +0.12
    -0.16
    459
    +0.12
    0.22
    Negative Logits
     constitutional
    -0.51
     Political
    -0.48
     Constitutional
    -0.47
     Rodham
    -0.46
     âĢİ
    -0.46
    iberal
    -0.45
     nonviolent
    -0.44
    âĢİ
    -0.42
     Osama
    -0.41
    osate
    -0.41
    POSITIVE LOGITS
     kinda
    0.47
     devs
    0.47
     usability
    0.45
     customization
    0.45
     backend
    0.44
     somew
    0.44
     haha
    0.43
     setups
    0.42
     optimization
    0.42
     presets
    0.41
    Activations Density 14.449%

    No Known Activations