INDEX

Explanations

phrases indicating agreement or compliance with a suggestion or idea

oai_token-act-pair · gpt-3.5-turbo

the phrase "go along," indicating compliance or agreement with something

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

ĻĤ

-0.70

ptive

-0.68

ĳå£«

-0.68

dinand

-0.66

etheus

-0.64

inction

-0.64

ilies

-0.63

orians

-0.63

iled

-0.63

cluding

-0.63

POSITIVE LOGITS

wagon

0.82

ments

0.70

 wagon

0.69

ities

0.67

 Vest

0.66

side

0.66

 along

0.65

agon

0.64

 bandwagon

0.62

 mate

0.60

Activations Density 0.014%

phrases indicating agreement or compliance with a suggestion or idea

the phrase "go along," indicating compliance or agreement with something

No Comments

No Known Activations