INDEX

Explanations

mentions of familiarity or lack thereof with certain topics or concepts

oai_token-act-pair · gpt-3.5-turbo

references to familiarity and unfamiliarity with concepts or subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

Ħ¢

-0.69

tein

-0.65

avorite

-0.63

 practicable

-0.63

ĺħ

-0.58

ument

-0.57

²¾

-0.57

©¶æ

-0.56

efficiency

-0.56

secondary

-0.56

POSITIVE LOGITS

izing

1.21

ized

1.19

ize

1.09

ising

1.06

ity

1.05

ities

1.02

ised

1.02

enough

1.00

izes

0.97

lly

0.95

Activations Density 0.032%

mentions of familiarity or lack thereof with certain topics or concepts

references to familiarity and unfamiliarity with concepts or subjects

No Comments

No Known Activations