INDEX

Explanations

phrases related to work or employment

oai_token-act-pair · gpt-3.5-turbo

sentences that imply or suggest strong negative consequences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

utical

-0.70

 commissions

-0.60

 representation

-0.59

 equip

-0.59

 Gregory

-0.57

xual

-0.57

 dissu

-0.56

 consideration

-0.56

cules

-0.56

 Versus

-0.56

POSITIVE LOGITS

ÂłÂłÂłÂłÂłÂłÂłÂł

1.02

WASHINGTON

0.90

Scroll

0.88

Reuters

0.84

SHARE

0.83

ÂłÂłÂłÂł

0.82

SPONSORED

0.81

Asked

0.80

CVE

0.80

SEE

0.79

Activations Density 0.405%

phrases related to work or employment

sentences that imply or suggest strong negative consequences

No Comments

No Known Activations