INDEX

Explanations

words related to achieving success or reaching specific outcomes

oai_token-act-pair · gpt-3.5-turbo

repeated references to quantity and measurement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

NetMessage

-0.74

etheless

-0.73

£ı

-0.65

Ô

-0.65

eatures

-0.65

agall

-0.65

 earnest

-0.64

lished

-0.63

ï¸

-0.63

Mn

-0.62

POSITIVE LOGITS

horn

0.70

 Fest

0.68

 regist

0.68

 pack

0.67

 holder

0.66

utils

0.62

icult

0.62

nance

0.61

otype

0.61

boards

0.61

Activations Density 0.806%

words related to achieving success or reaching specific outcomes

repeated references to quantity and measurement

No Comments

No Known Activations