INDEX

Explanations

phrases related to things that only appear to be a certain way

oai_token-act-pair · gpt-3.5-turbo

the word "seemingly" and its variations, indicating uncertainty or subjective perception

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

enter

-0.74

iaries

-0.68

ect

-0.62

jew

-0.61

winter

-0.61

ector

-0.60

 Excellence

-0.59

atin

-0.59

ther

-0.59

aches

-0.59

POSITIVE LOGITS

 seemingly

3.06

 seeming

1.90

 supposedly

1.90

 ostensibly

1.84

 purportedly

1.70

 apparently

1.65

 inexpl

1.55

 presumably

1.53

 mysteriously

1.49

 purported

1.39

Activations Density 0.018%

phrases related to things that only appear to be a certain way

the word "seemingly" and its variations, indicating uncertainty or subjective perception

No Comments

No Known Activations