INDEX

Explanations

phrases indicating a strong emotional reaction or emphasis

oai_token-act-pair · gpt-3.5-turbo

expressions of comparison or similarity

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

alogue

-0.75

hiba

-0.73

arers

-0.72

atform

-0.70

tein

-0.70

Versions

-0.69

FL

-0.68

Lower

-0.67

ircraft

-0.67

verning

-0.66

POSITIVE LOGITS

lihood

0.94

lier

0.80

liest

0.77

 crazy

0.77

wow

0.70

liness

0.69

ably

0.68

 goddamn

0.65

 parity

0.65

 crap

0.64

Activations Density 0.069%

phrases indicating a strong emotional reaction or emphasis

expressions of comparison or similarity

No Comments

No Known Activations