INDEX

Explanations

words that are part of the pattern "wn" followed by a number

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "own" and its various forms in different 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

uate

-0.91

itol

-0.72

ptive

-0.71

 infl

-0.67

 rehab

-0.64

orescent

-0.64

 recapt

-0.64

ttle

-0.63

uated

-0.62

 ions

-0.61

POSITIVE LOGITS

wn

1.09

iak

0.96

nesday

0.96

estern

0.96

elcome

0.95

erness

0.94

akening

0.92

sylvania

0.91

itness

0.88

akens

0.85

Activations Density 0.015%

words that are part of the pattern "wn" followed by a number

occurrences of the word "own" and its various forms in different contexts

No Comments

No Known Activations