INDEX

Explanations

numeric quantities with a comparative "than" followed by a number or specific unit of measurement

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate comparisons and statistics related to quantities and time

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

 brakes

-0.69

Afee

-0.56

apego

-0.55

 hesitated

-0.53

escription

-0.53

ÃįÃį

-0.52

 Russians

-0.52

hran

-0.52

 Ukrainians

-0.51

obyl

-0.51

POSITIVE LOGITS

 guise

0.81

 nutshell

0.70

 contexts

0.68

 vicinity

0.68

 realm

0.65

 manner

0.65

 precincts

0.64

vironments

0.63

 territory

0.63

 mode

0.61

Activations Density 1.496%

numeric quantities with a comparative "than" followed by a number or specific unit of measurement

phrases that indicate comparisons and statistics related to quantities and time

No Comments

No Known Activations