INDEX

Explanations

commands or actions related to enforcing rules and consequences, such as removing, terminating, banning, and disbanding

oai_token-act-pair · gpt-3.5-turbo

words and phrases related to negative outcomes or penalties

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

ricanes

-0.68

 lately

-0.58

inventoryQuantity

-0.56

 involves

-0.54

emis

-0.54

quartered

-0.53

ricane

-0.53

©¶æ

-0.52

onut

-0.52

 plurality

-0.52

POSITIVE LOGITS

 forever

1.24

 accordingly

1.01

 instantly

0.98

 automatically

0.98

.(

0.94

0.91

.[

0.91

 unless

0.90

.</

0.90

 regardless

0.90

Activations Density 0.557%

commands or actions related to enforcing rules and consequences, such as removing, terminating, banning, and disbanding

words and phrases related to negative outcomes or penalties

No Comments

No Known Activations