INDEX

Explanations

phrases related to a wide range of topics and subjects

oai_token-act-pair · gpt-3.5-turbo

articles and quantifiers used to describe various categories and scopes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 advis

-0.90

 occurrences

-0.84

 appointments

-0.82

 aliases

-0.81

 amendments

-0.80

 words

-0.79

 excuses

-0.79

 incidents

-0.78

 remarks

-0.78

 issues

-0.78

POSITIVE LOGITS

 variety

1.28

uras

1.25

 multitude

1.20

 plethora

1.16

 subset

1.07

 myriad

1.05

 wide

1.02

versive

1.02

 range

0.98

 wider

0.97

Activations Density 0.384%

phrases related to a wide range of topics and subjects

articles and quantifiers used to describe various categories and scopes

No Comments

No Known Activations