Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 17566
    Prev
    Next
    INDEX
    Explanations

    words related to perpetuating or maintaining specific beliefs, actions, or conditions

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "perpetuate" and its variations, indicating discussions about the continuation of certain situations or narratives

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ionic
    -0.90
    leaf
    -0.73
    kin
    -0.70
    atch
    -0.69
    ank
    -0.69
    arks
    -0.67
    anka
    -0.66
    rain
    -0.66
    iy
    -0.66
    ionics
    -0.64
    POSITIVE LOGITS
     perpet
    3.93
     perpetuate
    2.26
     perpetrated
    1.61
     exacerb
    1.33
     glor
    1.14
     legitim
    1.12
     perpetual
    1.10
     propag
    1.06
     deval
    1.04
     prolong
    1.03
    Activations Density 0.028%

    No Known Activations