INDEX

Explanations

pronouns and interrogative words indicating identification or questioning

oai_token-act-pair · gpt-3.5-turbo

pronouns related to collective identity or community involvement

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

 Increases

-0.63

 CITY

-0.60

 Cabin

-0.58

 Hunting

-0.58

 ACTIONS

-0.58

 Cancel

-0.57

 existed

-0.57

umar

-0.56

 Samp

-0.56

 Virus

-0.56

POSITIVE LOGITS

aning

0.81

 supposed

0.79

 kidding

0.78

avers

0.76

zzo

0.76

arth

0.76

soDeliveryDate

0.75

 aware

0.74

atics

0.72

omorphic

0.72

Activations Density 0.075%

pronouns and interrogative words indicating identification or questioning

pronouns related to collective identity or community involvement

No Comments

No Known Activations