© 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. 19541
    Prev
    Next
    INDEX
    Explanations

    the word "Med" or words related to it such as "Medellin"

    oai_token-act-pair · gpt-3.5-turbo

    mentions of the "Medellín Cartel."

    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
    chnology
    -0.81
     innocence
    -0.80
    ãģį
    -0.79
    ģĸ
    -0.72
    !/
    -0.70
    AUT
    -0.69
    RECT
    -0.67
    holding
    -0.66
    ngth
    -0.66
    cipline
    -0.66
    POSITIVE LOGITS
    iterranean
    1.14
    ications
    1.12
    allion
    1.01
    itations
    0.99
    ient
    0.98
    iator
    0.98
    ril
    0.95
    itate
    0.94
    ician
    0.94
    iant
    0.94
    Activations Density 0.009%

    No Known Activations