INDEX

Explanations

specific instances or examples of something being described or mentioned

oai_token-act-pair · gpt-3.5-turbo

references to specific items or entities referred to as "ones."

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

SEA

-0.73

 Falcons

-0.71

 Membership

-0.65

Leader

-0.64

Rush

-0.63

ORY

-0.62

osponsors

-0.62

 Nanto

-0.62

Attorney

-0.61

Definition

-0.61

POSITIVE LOGITS

hots

1.17

omething

0.91

hot

0.90

selves

0.80

 eyed

0.79

creen

0.78

ettings

0.77

uits

0.75

elf

0.75

cale

0.74

Activations Density 0.024%

specific instances or examples of something being described or mentioned

references to specific items or entities referred to as "ones."

No Comments

No Known Activations