INDEX

Explanations

financial terms

oai_token-act-pair · gpt-3.5-turbo

phrases that describe items or entities being categorized or documented in lists

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

qua

-0.76

issance

-0.72

assi

-0.71

ggles

-0.69

ework

-0.69

selves

-0.64

ipal

-0.64

Gaw

-0.64

scape

-0.63

perty

-0.62

POSITIVE LOGITS

 listing

1.07

 listings

1.00

 listed

0.88

erv

0.80

 lists

0.77

 directories

0.75

 Lists

0.74

 below

0.72

 above

0.70

 widget

0.70

Activations Density 0.014%

financial terms

phrases that describe items or entities being categorized or documented in lists

No Comments

No Known Activations