INDEX

Explanations

words related to decreases or declines

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "in" as an indicator of contextual positioning

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

IDE

-0.73

arers

-0.68

soever

-0.67

NOW

-0.67

Nat

-0.67

Registered

-0.67

çīĪ

-0.64

linger

-0.64

pecially

-0.63

 endors

-0.63

POSITIVE LOGITS

effic

1.33

efficiency

1.32

ordinate

1.16

 relation

1.00

 situ

0.99

animate

0.99

activity

0.98

 favor

0.97

 conjunction

0.97

 accordance

0.97

Activations Density 0.256%

words related to decreases or declines

occurrences of the word "in" as an indicator of contextual positioning

No Comments

No Known Activations