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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 8-RES-JB
    11. 6071
    Prev
    Next
    INDEX
    Explanations

    words related to the concept of squeezing or pressing with force

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "squish" or variations thereof

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Luther
    -0.70
     reconciliation
    -0.69
     dating
    -0.66
     masculinity
    -0.66
     neglect
    -0.65
    urally
    -0.64
     Archdemon
    -0.60
    WAYS
    -0.59
     rested
    -0.59
     assum
    -0.58
    POSITIVE LOGITS
    irts
    1.21
    atters
    1.17
    ishable
    1.16
    alid
    1.14
    irting
    1.13
    irrel
    1.12
    awk
    1.09
    atter
    1.08
    atted
    1.06
    ishy
    1.05
    Activations Density 0.024%

    No Known Activations