INDEX

Explanations

instances referring to being 'in' a certain situation or position

oai_token-act-pair · gpt-3.5-turbo

occurrences of the phrase "in" followed by various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Malays

-0.66

 Detected

-0.66

 notwithstanding

-0.66

Published

-0.64

llan

-0.60

 arising

-0.60

 vanquished

-0.57

 tallied

-0.56

 aside

-0.56

 intermitt

-0.55

POSITIVE LOGITS

offensive

1.06

animate

1.06

ordinate

1.05

humane

1.01

ching

0.98

authent

0.96

clusively

0.90

escap

0.89

strument

0.88

patient

0.88

Activations Density 0.181%

instances referring to being 'in' a certain situation or position

occurrences of the phrase "in" followed by various contexts

No Comments

No Known Activations