INDEX

Explanations

mentions of government officials, particularly prime ministers and ministers

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

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

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Locations

-0.77

 Condition

-0.73

 Blend

-0.70

edIn

-0.70

 Latter

-0.70

 Warfare

-0.67

 Casting

-0.67

Inc

-0.66

 Contents

-0.65

 Conditions

-0.64

POSITIVE LOGITS

's

0.92

 himself

0.88

 tasked

0.86

 office

0.85

 vetoed

0.85

ÃŃs

0.84

 apologised

0.83

 congratulated

0.83

 overseeing

0.82

 apologized

0.82

Activations Density 0.160%

mentions of government officials, particularly prime ministers and ministers

No Comments

No Known Activations