INDEX

Explanations

hyphenated words ending with 'to' or 'between to'

oai_token-act-pair · gpt-3.5-turbo

phrases related to negative experiences or conditions

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

 Seym

-0.80

 aborted

-0.65

 coats

-0.64

 McKin

-0.64

 Viper

-0.63

 questioned

-0.61

 Tempest

-0.61

pta

-0.59

 Rolls

-0.58

 Dickinson

-0.58

POSITIVE LOGITS

advertising

0.92

elect

0.90

date

0.89

Egypt

0.85

morrow

0.84

nom

0.84

reach

0.84

equ

0.83

earth

0.83

arms

0.83

Activations Density 0.030%

hyphenated words ending with 'to' or 'between to'

phrases related to negative experiences or conditions

No Comments

No Known Activations