INDEX

Explanations

phrases related to the concept of "higher"

oai_token-act-pair · gpt-3.5-turbo

references to elevated levels or standards in various contexts

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

Sho

-0.66

 Paraly

-0.66

ä¸ī

-0.65

pty

-0.64

 Scythe

-0.62

Reply

-0.62

Ring

-0.61

icate

-0.61

 Canaver

-0.61

bers

-0.60

POSITIVE LOGITS

 than

1.05

than

1.03

iating

0.89

 tiers

0.88

 pitched

0.85

 levels

0.85

 referen

0.81

 priced

0.81

iated

0.81

 earners

0.79

Activations Density 0.035%

phrases related to the concept of "higher"

references to elevated levels or standards in various contexts

No Comments

No Known Activations