INDEX

Explanations

phrases related to achievement or accomplishment

oai_token-act-pair · gpt-3.5-turbo

instances of the word "made."

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

 conditioning

-0.69

externalActionCode

-0.67

 stripe

-0.67

 Adin

-0.64

heter

-0.64

Mub

-0.61

 heartbeat

-0.61

 ILCS

-0.60

 fray

-0.58

iology

-0.58

POSITIVE LOGITS

netflix

0.91

 sure

0.90

roid

0.87

awaru

0.80

emort

0.80

itions

0.79

iversal

0.77

ÄŁ

0.77

dinand

0.76

 landfall

0.75

Activations Density 0.064%

phrases related to achievement or accomplishment

instances of the word "made."

No Comments

No Known Activations