INDEX

Explanations

terms related to covering costs or expenses

oai_token-act-pair · gpt-3.5-turbo

instances of the word "cover" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Sabha

-0.94

rous

-0.75

arb

-0.67

jet

-0.66

rouse

-0.65

urt

-0.64

 fortune

-0.63

rained

-0.63

lua

-0.61

vil

-0.61

POSITIVE LOGITS

iday

0.86

ages

0.82

utical

0.81

alls

0.79

Cover

0.78

krit

0.77

 cover

0.76

ãĥ¯

0.75

 covering

0.75

 covers

0.73

Activations Density 0.030%

terms related to covering costs or expenses

instances of the word "cover" in various contexts

No Comments

No Known Activations