© 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. 39634
    Prev
    Next
    INDEX
    Explanations

    units of measure and weights, such as pounds, kilograms, and gallons, as well as food items and meal preparations

    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
    55
    +0.06
    0.5%
    320
    +0.05
    0.5%
    158
    +0.05
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    442
    +0.06
    0.07
    46
    +0.05
    0.06
    582
    +0.05
    0.02
    Negative Logits
    PsyNetMessage
    -0.40
     Sind
    -0.33
     Telecommunications
    -0.33
    911
    -0.33
     Initi
    -0.32
     Garmin
    -0.32
     Rite
    -0.32
    TION
    -0.32
     Gat
    -0.32
     Journalists
    -0.31
    POSITIVE LOGITS
    cot
    0.51
    arella
    0.45
    flies
    0.44
    jee
    0.44
    ergic
    0.44
    assy
    0.43
    mouse
    0.42
    fly
    0.42
    yang
    0.42
    ucker
    0.41
    Activations Density 15.841%

    No Known Activations