INDEX

Explanations

the term "used to" followed by a verb phrase, indicating familiarity or habituation with a particular action

oai_token-act-pair · gpt-3.5-turbo

instances of the word "used" in various contexts related to adaptation and familiarity

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

tein

-0.69

etsk

-0.61

 termination

-0.58

 commencement

-0.56

hemy

-0.55

 State

-0.55

 enterprise

-0.54

 feasibility

-0.54

 deviation

-0.54

 obtained

-0.53

POSITIVE LOGITS

bodied

0.88

ivated

0.85

ience

0.83

 aback

0.82

ivating

0.81

isp

0.79

thumbnails

0.78

icip

0.75

ickle

0.75

ibly

0.75

Activations Density 0.096%

the term "used to" followed by a verb phrase, indicating familiarity or habituation with a particular action

instances of the word "used" in various contexts related to adaptation and familiarity

No Comments

No Known Activations