INDEX

Explanations

words related to comparison, evaluation, or emphasis

oai_token-act-pair · gpt-3.5-turbo

words that indicate emphasis or intensity in statements

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

å£«

-0.68

available

-0.61

anium

-0.60

cream

-0.59

taboola

-0.58

Micro

-0.56

enium

-0.55

unta

-0.55

etus

-0.55

worker

-0.55

POSITIVE LOGITS

 unsurprisingly

0.55

 Woodward

0.53

 Schiff

0.53

 McCull

0.52

 Miliband

0.51

 Harriet

0.50

 Starr

0.50

 Sharif

0.49

 Rosenberg

0.49

 unsuccessfully

0.49

Activations Density 0.942%

words related to comparison, evaluation, or emphasis

words that indicate emphasis or intensity in statements

No Comments

No Known Activations