INDEX

Explanations

code snippets with logical conditions

oai_token-act-pair · gpt-3.5-turbo

the conjunction "and" used 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

meet

-0.73

Gaza

-0.72

actionDate

-0.69

--------------------------------------------------------

-0.66

fleet

-0.63

ONSORED

-0.62

OTA

-0.61

lift

-0.60

graduate

-0.59

çīĪ

-0.59

POSITIVE LOGITS

 therefore

1.31

 thus

1.22

 consequently

1.19

 hence

1.19

 then

1.10

 THEN

1.09

 thereby

0.97

 furthermore

0.95

 secondly

0.93

rogens

0.90

Activations Density 0.548%

code snippets with logical conditions

the conjunction "and" used in various contexts

No Comments

No Known Activations