INDEX

Explanations

words related to regulations being rolled back

oai_token-act-pair · gpt-3.5-turbo

phrases related to reversing policies or actions

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

legged

-0.70

Ã¡n

-0.67

ussen

-0.66

Warning

-0.66

wyn

-0.65

oled

-0.64

raq

-0.64

acht

-0.64

Eye

-0.63

SELECT

-0.62

POSITIVE LOGITS

 globalization

0.72

ively

0.66

 disco

0.65

 misunderstand

0.64

 Prohibition

0.63

 declines

0.61

 answering

0.60

theless

0.60

 overc

0.59

 interruption

0.59

Activations Density 0.228%

words related to regulations being rolled back

phrases related to reversing policies or actions

No Comments

No Known Activations