INDEX

Explanations

the word "Long" in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to "Long" 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 @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

enge

-0.74

enic

-0.72

awaru

-0.65

asse

-0.65

atis

-0.63

doms

-0.63

OTUS

-0.63

 recess

-0.62

atically

-0.62

 nonex

-0.62

POSITIVE LOGITS

 Long

3.57

Long

2.68

 LONG

2.17

long

1.89

 Short

1.76

 long

1.41

Short

1.28

short

1.25

 longest

1.18

 longer

1.17

Activations Density 0.011%

the word "Long" in various contexts

references to "Long" in various contexts

No Comments

No Known Activations