INDEX

Explanations

phrases indicating personal insight or knowledge

oai_token-act-pair · gpt-3.5-turbo

phrases that emphasize understanding or acknowledgment

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

utenberg

-0.66

aez

-0.62

anmar

-0.61

 recomm

-0.60

arthed

-0.58

ossibility

-0.57

omal

-0.55

acco

-0.55

 fortun

-0.55

mination

-0.55

POSITIVE LOGITS

 what

1.08

why

1.00

how

1.00

terday

0.97

 WHAT

0.88

 whats

0.85

what

0.81

 exactly

0.75

lege

0.71

 somet

0.70

Activations Density 0.034%

phrases indicating personal insight or knowledge

phrases that emphasize understanding or acknowledgment

No Comments

No Known Activations