INDEX

Explanations

questions starting with "What do you think?" or "What does this mean?"

oai_token-act-pair · gpt-3.5-turbo

questions seeking opinions or clarification

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

WAYS

-0.82

ahime

-0.69

 Bridges

-0.67

gra

-0.66

legram

-0.65

Lago

-0.64

fox

-0.63

iza

-0.63

gart

-0.62

knit

-0.61

POSITIVE LOGITS

?]

0.81

 transpired

0.72

ãĤ»

0.71

 wrong

0.67

?)

0.67

 glean

0.66

 happen

0.65

?!"

0.65

?!

0.64

 characterize

0.64

Activations Density 0.050%

questions starting with "What do you think?" or "What does this mean?"

questions seeking opinions or clarification

No Comments

No Known Activations