INDEX

Explanations

keyword 'be'

oai_token-act-pair · gpt-3.5-turbo

the verb "be" in various forms and contexts

oai_token-act-pair · gpt-4o-mini Triggered 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

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Bars

-0.74

ciating

-0.67

terday

-0.66

might

-0.61

ocre

-0.61

plex

-0.59

fortunately

-0.58

 compose

-0.57

 Bend

-0.57

strous

-0.57

POSITIVE LOGITS

 able

1.05

heading

0.97

 judged

0.95

AMS

0.95

fall

0.95

 replaced

0.94

 rewarded

0.93

 subjected

0.92

 deemed

0.92

fitting

0.91

Activations Density 0.230%

keyword 'be'

the verb "be" in various forms and contexts

No Comments

No Known Activations