© 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. 0-RES-JB
    11. 23172
    Prev
    Next
    INDEX
    Explanations

    expressions of strong negative emotions, particularly outrage

    oai_token-act-pair · gpt-3.5-turbo

    expressions of public outrage

    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
    alone
    -0.70
     Selected
    -0.70
    orporated
    -0.68
    Stud
    -0.68
    chie
    -0.68
    prints
    -0.65
     Formation
    -0.64
    aka
    -0.63
     Strategies
    -0.63
     Physical
    -0.63
    POSITIVE LOGITS
     outrage
    3.48
     indignation
    2.23
     outcry
    2.13
     uproar
    2.08
     outraged
    1.82
     anger
    1.70
     disgust
    1.60
     fury
    1.59
     condemnation
    1.48
     rage
    1.44
    Activations Density 0.020%

    No Known Activations