INDEX

Explanations

evidence of actions or events

oai_token-act-pair · gpt-3.5-turbo

action words that suggest activity or engagement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

VIDIA

-0.66

 notor

-0.65

ilitary

-0.64

avascript

-0.62

 encount

-0.62

 Palestin

-0.62

ccording

-0.60

 reluct

-0.60

theless

-0.60

 conflic

-0.59

POSITIVE LOGITS

ings

1.28

able

1.16

ingly

1.12

ables

1.09

backs

0.98

ments

0.97

INGS

0.91

ably

0.91

ability

0.90

downs

0.84

Activations Density 0.757%

evidence of actions or events

action words that suggest activity or engagement

No Comments

No Known Activations