INDEX

Explanations

phrases expressing acceptance of responsibility or specific conditions

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

phrases related to acceptance and responsibility

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

vertisement

-0.85

etheus

-0.74

fox

-0.73

obbies

-0.72

ionage

-0.71

iferation

-0.70

 Indust

-0.69

Scope

-0.67

APH

-0.67

eches

-0.66

POSITIVE LOGITS

 responsibility

1.34

 bribes

1.08

 invitations

1.02

 resignation

1.00

 invitation

0.99

 defeat

0.91

 apologies

0.91

 donations

0.88

 payment

0.88

 accept

0.88

Activations Density 0.162%

phrases expressing acceptance of responsibility or specific conditions

phrases related to acceptance and responsibility

No Comments

No Known Activations