INDEX

Explanations

questions starting with "How"

oai_token-act-pair · gpt-3.5-turbo

questions that begin with "How"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

-0.74

Ãį

-0.66

iculture

-0.62

agonists

-0.61

ceptions

-0.61

 grounds

-0.61

yak

-0.60

 fide

-0.60

 territory

-0.58

uthor

-0.58

POSITIVE LOGITS

soever

1.11

beit

0.90

ever

0.90

HCR

0.82

ling

0.76

leep

0.76

ls

0.75

ells

0.74

itzer

0.72

bill

0.70

Activations Density 0.057%

questions starting with "How"

questions that begin with "How"

No Comments

No Known Activations