INDEX

Explanations

unusual or surprising events or situations

oai_token-act-pair · gpt-3.5-turbo

references to being unheard or overlooked

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

unks

-0.70

 Flavoring

-0.69

apa

-0.68

 Hilton

-0.67

olars

-0.65

idate

-0.64

 drainage

-0.64

ouf

-0.62

addy

-0.61

olar

-0.61

POSITIVE LOGITS

 unheard

1.27

=]

0.87

entimes

0.80

Ĳ

0.77

DIV

0.76

lihood

0.74

ICLE

0.71

Write

0.71

ĭ

0.70

 surpr

0.70

Activations Density 0.012%

unusual or surprising events or situations

references to being unheard or overlooked

No Comments

No Known Activations