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. 8-RES-JB
    6. 21383
    Prev
    Next
    INDEX
    Explanations

    phrases related to politics, conflicts, and controversial events

    oai_token-act-pair · gpt-3.5-turbo

    topics related to political and social issues, particularly focusing on conflict and injustice

    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
    ographer
    -0.78
    ograph
    -0.74
    ationally
    -0.70
    rb
    -0.68
    ographers
    -0.65
    ivist
    -0.65
    ographically
    -0.63
    ocrat
    -0.62
    eworks
    -0.62
    Standard
    -0.62
    POSITIVE LOGITS
     inability
    1.30
     refusal
    1.20
     unwillingness
    1.13
     lack
    1.12
     reluctance
    1.05
     mishand
    1.01
     abandonment
    1.00
     willingness
    1.00
     misunderstanding
    0.97
     insistence
    0.97
    Activations Density 1.253%

    No Known Activations