INDEX

Explanations

words related to organizing or structuring data

oai_token-act-pair · gpt-3.5-turbo

variations of the word "arrange"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ĨĴ

-0.84

ku

-0.70

 Pixie

-0.70

fulness

-0.69

utral

-0.68

manship

-0.65

————

-0.65

google

-0.64

¥µ

-0.64

 Europa

-0.63

POSITIVE LOGITS

ays

1.03

agher

1.02

arr

0.99

abis

0.97

interstitial

0.94

iot

0.93

andom

0.92

abilia

0.92

agh

0.90

ader

0.89

Activations Density 0.005%

words related to organizing or structuring data

variations of the word "arrange"

No Comments

No Known Activations