INDEX

Explanations

phrases related to approval or authorization

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate standards or norms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Disciple

-0.63

ierre

-0.61

HAEL

-0.59

 Pend

-0.59

 srfAttach

-0.58

 Dickinson

-0.58

 Drivers

-0.58

 Divinity

-0.57

ATTLE

-0.56

 Vaughn

-0.56

POSITIVE LOGITS

ahead

1.17

bet

0.93

ither

0.90

gettable

0.88

between

0.86

around

0.86

along

0.83

too

0.81

round

0.80

cart

0.80

Activations Density 0.034%

phrases related to approval or authorization

phrases that indicate standards or norms

No Comments

No Known Activations