INDEX

Explanations

questions, particularly those ending with a question mark and focusing on decision-making

oai_token-act-pair · gpt-3.5-turbo

rhetorical questions and expressions of uncertainty

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

oun

-0.56

hemer

-0.56

ogly

-0.53

bryce

-0.53

utor

-0.52

yp

-0.52

ouf

-0.52

DX

-0.51

outhern

-0.50

ogyn

-0.50

POSITIVE LOGITS

 Eventually

0.68

 Especially

0.63

 Ultimately

0.62

 Whether

0.61

 regardless

0.61

 lest

0.61

 Otherwise

0.60

 Instead

0.60

 Unless

0.58

 inaction

0.57

Activations Density 1.064%

questions, particularly those ending with a question mark and focusing on decision-making

rhetorical questions and expressions of uncertainty

No Comments

No Known Activations