INDEX

Explanations

the word "strain"

oai_token-act-pair · gpt-4-turbo

instances of the word "strain"

oai_token-act-pair · gpt-3.5-turbo

references to different types of strains, particularly in medical or biological contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

*/(

-1.00

swick

-0.88

izons

-0.80

nect

-0.80

vertis

-0.78

inez

-0.76

vernment

-0.75

emouth

-0.73

fman

-0.73

lymp

-0.73

POSITIVE LOGITS

 strain

1.27

 strains

1.25

 spores

0.77

weed

0.73

 bred

0.73

 Vaugh

0.72

encies

0.70

 thereof

0.69

 infections

0.67

enza

0.64

Activations Density 0.008%

the word "strain"

instances of the word "strain"

references to different types of strains, particularly in medical or biological contexts

No Comments

No Known Activations