INDEX

Explanations

words ending in the letter 's'

oai_token-act-pair · gpt-3.5-turbo

instances of the letter 's' in various contexts

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

 satellite

-0.68

eh

-0.68

è»

-0.67

 fuss

-0.65

oct

-0.61

onyms

-0.60

ãĤ¿

-0.57

 fatalities

-0.57

eur

-0.56

 hect

-0.55

POSITIVE LOGITS

ourcing

0.96

ources

0.94

hips

0.91

linger

0.90

uddenly

0.89

wered

0.89

borough

0.89

addle

0.87

olute

0.86

atisf

0.84

Activations Density 0.088%

words ending in the letter 's'

instances of the letter 's' in various contexts

No Comments

No Known Activations