© 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. 4-RES-JB
    11. 10168
    Prev
    Next
    INDEX
    Explanations

    words related to negative qualities or conditions, such as maladies or poor performance

    oai_token-act-pair · gpt-3.5-turbo

    negative descriptors relating to quality or performance

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

    No Comments

    Negative Logits
    oother
    -0.70
    ourage
    -0.66
    omet
    -0.66
    arij
    -0.62
    ju
    -0.62
     lat
    -0.61
     congr
    -0.60
    andowski
    -0.59
     COURT
    -0.58
    ilon
    -0.57
    POSITIVE LOGITS
    bilt
    0.83
    luster
    0.80
     miser
    0.78
    theless
    0.76
    ocre
    0.74
     underest
    0.73
    ishly
    0.73
     underestimate
    0.71
    butt
    0.71
     disappoint
    0.70
    Activations Density 0.114%

    No Known Activations