INDEX

Explanations

mentions of being sick or health-related descriptions

oai_token-act-pair · gpt-3.5-turbo

repeated references to the term "sick" and its related contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 unlaw

-0.67

 distingu

-0.66

CLS

-0.66

 Unch

-0.65

rul

-0.65

 laun

-0.64

wcsstore

-0.63

 unden

-0.62

 Hier

-0.61

 Decoder

-0.60

POSITIVE LOGITS

ening

1.61

ened

1.54

bay

1.23

er

1.10

nesses

1.02

le

0.99

ly

0.99

ens

0.98

est

0.97

igans

0.94

Activations Density 0.028%

mentions of being sick or health-related descriptions

repeated references to the term "sick" and its related contexts

No Comments

No Known Activations