INDEX

Explanations

words related to negation or negative contexts

oai_token-act-pair · gpt-3.5-turbo

negative concepts or terms related to negativity and rejection

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ORGE

-0.82

BuyableInstoreAndOnline

-0.77

Â¯Â¯

-0.76

DragonMagazine

-0.76

Bloom

-0.76

realDonaldTrump

-0.74

OPLE

-0.74

WHERE

-0.72

strap

-0.72

OHN

-0.72

POSITIVE LOGITS

otiation

1.41

oti

1.35

atives

1.14

neg

1.08

rito

0.92

lect

0.89

ativity

0.88

lected

0.86

atively

0.83

isions

0.82

Activations Density 0.011%

words related to negation or negative contexts

negative concepts or terms related to negativity and rejection

No Comments

No Known Activations