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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 11-RES-JB
    6. 9771
    Prev
    Next
    INDEX
    Explanations

    decisions or actions indicating a negative choice or outcome

    oai_token-act-pair · gpt-3.5-turbo

    negative expressions or phrases indicating opposition or rejection

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    verbs
    -0.71
    initialized
    -0.67
     dashed
    -0.67
    perfect
    -0.67
    Weak
    -0.67
     illiter
    -0.64
    \\\\\\\\
    -0.64
     Basics
    -0.63
    Perfect
    -0.62
     debunked
    -0.62
    POSITIVE LOGITS
     renew
    1.46
     bother
    1.27
     pursue
    1.21
     participate
    1.14
     anymore
    1.04
     proceed
    1.04
     attend
    1.02
     continue
    0.98
     bud
    0.96
     tolerate
    0.96
    Activations Density 0.213%

    No Known Activations