INDEX

Explanations

the concept of something being at risk or in jeopardy

oai_token-act-pair · gpt-3.5-turbo

references to ownership or investment stakes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

pmwiki

-0.81

eric

-0.68

mph

-0.67

hammad

-0.61

ulia

-0.61

ysis

-0.60

eus

-0.60

ICA

-0.60

iferation

-0.60

nesota

-0.60

POSITIVE LOGITS

holder

1.41

 stake

1.18

pole

1.07

holders

0.94

 stakes

0.87

 srfAttach

0.77

rament

0.70

holding

0.69

lass

0.69

iffs

0.67

Activations Density 0.007%

the concept of something being at risk or in jeopardy

references to ownership or investment stakes

No Comments

No Known Activations