INDEX

Explanations

phrases related to status or classification

oai_token-act-pair · gpt-3.5-turbo

statements about the status or actions of various subjects

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

iasco

-0.73

adelphia

-0.67

cess

-0.65

ansom

-0.63

raught

-0.63

olester

-0.62

ruct

-0.62

MSN

-0.61

oj

-0.60

 Liqu

-0.60

POSITIVE LOGITS

 consequently

1.12

 therefore

1.12

 accordingly

1.08

 thus

0.83

 furthermore

0.82

 thereafter

0.80

 subsequently

0.79

 thereby

0.77

 thereof

0.73

 liable

0.71

Activations Density 0.278%

phrases related to status or classification

statements about the status or actions of various subjects

No Comments

No Known Activations