© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    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. 978
    Prev
    Next
    INDEX
    Explanations

    the term "strings" used in a metaphorical or abstract sense for tying events, relationships, or themes together

    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
    167
    +0.09
    0.5%
    189
    +0.09
    0.5%
    587
    +0.08
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.09
    0.05
    288
    +0.09
    0.06
    129
    +0.08
    0.03
    Negative Logits
    alach
    -0.62
    edia
    -0.60
    rompt
    -0.56
    apa
    -0.56
     deserts
    -0.56
     Scotia
    -0.54
     Agenda
    -0.53
     Trevor
    -0.53
    iator
    -0.53
     Maw
    -0.52
    POSITIVE LOGITS
    angle
    0.65
     percussion
    0.63
     horns
    0.63
     tighten
    0.63
    angled
    0.62
     necks
    0.61
    heads
    0.59
    ouls
    0.59
    eners
    0.59
    picked
    0.58
    Activations Density 0.489%

    No Known Activations