INDEX

Explanations

references to geographical locations, specifically in relation to political, administrative, or crisis-related contexts

oai_token-act-pair · gpt-3.5-turbo

the conjunction "and" used in various contexts

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

DragonMagazine

-0.84

BIL

-0.82

Sov

-0.78

lash

-0.76

JJ

-0.69

uper

-0.69

cel

-0.69

HK

-0.68

LOCK

-0.68

TP

-0.67

POSITIVE LOGITS

 thereby

0.94

 hence

0.93

 thus

0.92

 thereafter

0.91

 consequently

0.91

 vicinity

0.89

 thence

0.88

 vice

0.86

 elsewhere

0.86

 oversaw

0.81

Activations Density 0.509%

references to geographical locations, specifically in relation to political, administrative, or crisis-related contexts

the conjunction "and" used in various contexts

No Comments

No Known Activations