INDEX

Explanations

phrases expressing memory or lasting impact

oai_token-act-pair · gpt-3.5-turbo

expressions of memory and forgetfulness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 liner

-0.68

catentry

-0.62

NSA

-0.60

 portfolios

-0.60

 antioxid

-0.59

Dialogue

-0.58

iHUD

-0.56

Alert

-0.54

 shoulders

-0.53

still

-0.53

POSITIVE LOGITS

nor

1.01

 slightest

0.90

 whatsoever

0.83

rences

0.80

 anymore

0.76

 dime

0.74

 dreamed

0.73

apon

0.72

 except

0.71

 bothered

0.70

Activations Density 0.223%

phrases expressing memory or lasting impact

expressions of memory and forgetfulness

No Comments

No Known Activations