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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlog/PodcastSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlog/PodcastGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 5-RES-JB
    6. 2392
    Prev
    Next
    INDEX
    Explanations

    statements about political decisions and actions

    oai_token-act-pair · gpt-3.5-turbo

    references to legislative or organizational actions and proposals related to health and social care

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 5-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    5
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    colored
    -0.96
    avored
    -0.93
     favors
    -0.93
     honors
    -0.90
    Defense
    -0.89
     coloring
    -0.89
     honor
    -0.89
    Los
    -0.84
    avior
    -0.82
     colors
    -0.82
    POSITIVE LOGITS
     recognise
    1.51
     realise
    1.45
     £
    1.43
     Labour
    1.38
     recognised
    1.37
     apologise
    1.34
     Ukip
    1.34
    £
    1.33
     emphas
    1.32
     organising
    1.32
    Activations Density 1.663%

    No Known Activations