INDEX

Explanations

words or phrases related to responses or reactions

oai_token-act-pair · gpt-3.5-turbo

phrases indicating answers or reactions to questions or situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

flo

-0.71

Ban

-0.70

illin

-0.67

ider

-0.65

onis

-0.64

 awoken

-0.63

utters

-0.63

anon

-0.62

vet

-0.61

tis

-0.61

POSITIVE LOGITS

 guise

0.97

 midst

0.80

 fashion

0.77

 context

0.76

 haste

0.74

 manner

0.70

 form

0.68

ItemTracker

0.68

 vicinity

0.66

ãĤ°

0.66

Activations Density 0.203%

words or phrases related to responses or reactions

phrases indicating answers or reactions to questions or situations

No Comments

No Known Activations