INDEX

Explanations

phrases related to personal experiences and emotions

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "in" and references to places or contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

-0.78

Redditor

-0.72

Released

-0.71

aic

-0.66

ä

-0.65

ãĤĴ

-0.64

avanaugh

-0.63

 Released

-0.62

erenn

-0.61

llan

-0.60

POSITIVE LOGITS

 terms

1.24

my

1.01

terms

0.92

laughs

0.90

 retrospect

0.89

 regards

0.88

 hindsight

0.87

effic

0.84

 particular

0.82

 fact

0.80

Activations Density 0.262%

phrases related to personal experiences and emotions

occurrences of the word "in" and references to places or contexts

No Comments

No Known Activations