INDEX

Explanations

phrases expressing feelings of reward, satisfaction, and emotional responses to experiences

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

legal

-0.64

Phill

-0.62

Newsletter

-0.60

Bridge

-0.59

Writer

-0.59

Roy

-0.58

 Ashton

-0.58

hawks

-0.58

 merger

-0.57

umar

-0.56

POSITIVE LOGITS

 yourself

1.16

 yourselves

0.98

 Yourself

0.83

 wasting

0.76

 temptation

0.74

 your

0.74

 wondering

0.73

ãĤ¦ãĤ¹

0.72

 wiser

0.71

 forgiven

0.70

Activations Density 0.264%

phrases expressing feelings of reward, satisfaction, and emotional responses to experiences

No Comments

No Known Activations