INDEX

Explanations

terms related to the early years of different contexts or people

oai_token-act-pair · gpt-3.5-turbo

references to the early 20th century

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

olulu

-0.80

ittee

-0.79

uyomi

-0.78

ortium

-0.77

retched

-0.75

laws

-0.75

edIn

-0.74

mbuds

-0.72

ingen

-0.72

channelAvailability

-0.71

POSITIVE LOGITS

 stages

0.93

 adop

0.92

 onset

0.85

 drafts

0.83

 prototypes

0.80

 phases

0.79

prototype

0.77

 childhood

0.75

 beginnings

0.75

-'

0.75

Activations Density 0.028%

terms related to the early years of different contexts or people

references to the early 20th century

No Comments

No Known Activations