INDEX

Explanations

words related to controversy and turmoil in geopolitical contexts

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Desk

-0.74

 guessing

-0.69

roth

-0.67

Ble

-0.62

river

-0.61

Tax

-0.58

ilet

-0.56

LY

-0.56

roy

-0.56

icularly

-0.56

POSITIVE LOGITS

 accompanies

1.15

 surrounds

1.07

 arose

1.01

 preceded

1.01

 comprise

0.94

 resulted

0.94

 surround

0.93

 governs

0.91

 existed

0.91

 separates

0.91

Activations Density 0.557%

words related to controversy and turmoil in geopolitical contexts

No Comments

No Known Activations