INDEX

Explanations

expressions related to understanding or getting something

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "get it."

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

izable

-0.62

 Mans

-0.60

avage

-0.60

hips

-0.58

 Friend

-0.57

currently

-0.57

Sund

-0.57

Lar

-0.55

ãĥ©ãĥ³

-0.54

CHO

-0.54

POSITIVE LOGITS

alian

1.19

chy

1.09

unes

0.95

iner

0.91

self

0.91

ueller

0.81

atic

0.79

geist

0.77

asca

0.75

atical

0.70

Activations Density 0.151%

expressions related to understanding or getting something

instances of the phrase "get it."

No Comments

No Known Activations