INDEX

Explanations

verbs ending in 'ing' or 'ed' indicating continuous or past actions

oai_token-act-pair · gpt-3.5-turbo

words that represent ongoing or completed actions, often in various grammatical forms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 Lauder

-0.73

 facult

-0.62

ORY

-0.62

itans

-0.61

OSP

-0.61

VPN

-0.58

rophe

-0.58

abet

-0.57

ITNESS

-0.57

MAT

-0.57

POSITIVE LOGITS

 away

0.87

tips

0.83

addle

0.81

ulate

0.80

 down

0.78

 Away

0.77

away

0.73

out

0.73

ishly

0.72

tails

0.72

Activations Density 0.106%

verbs ending in 'ing' or 'ed' indicating continuous or past actions

words that represent ongoing or completed actions, often in various grammatical forms

No Comments

No Known Activations