INDEX

Explanations

instances of time-related events described with the word 'just'

oai_token-act-pair · gpt-3.5-turbo

the word "just," indicating instances of immediacy or recent actions

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

xual

-0.74

 cous

-0.68

 confir

-0.68

 sacrific

-0.68

idon

-0.67

ixel

-0.66

 seiz

-0.65

 challeng

-0.65

aware

-0.64

undai

-0.64

POSITIVE LOGITS

ifications

1.14

ifiable

1.05

IFIC

0.90

itia

0.90

IFIED

0.89

if

0.88

ices

0.88

ICES

0.84

 WATCHED

0.83

ifi

0.80

Activations Density 0.100%

instances of time-related events described with the word 'just'

the word "just," indicating instances of immediacy or recent actions

No Comments

No Known Activations