INDEX

Explanations

phrases related to conjunctions

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "and" 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 @ 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

owered

-0.85

ãĥ³ãĤ¸

-0.78

hered

-0.78

ãĤ´ãĥ³

-0.78

oward

-0.73

zan

-0.67

pared

-0.67

osed

-0.67

held

-0.66

elle

-0.66

POSITIVE LOGITS

 secondly

1.18

 therefore

1.10

 moreover

1.06

 there

1.06

 consequently

1.05

 hence

1.04

 they

1.04

we

1.02

 furthermore

0.99

it

0.98

Activations Density 0.394%

phrases related to conjunctions

occurrences of the word "and" in various contexts

No Comments

No Known Activations