INDEX

Explanations

names, specifically last names

oai_token-act-pair · gpt-3.5-turbo

occurrences of the suffix 'ini' 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 @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

¥µ

-0.77

é¾įå¥ĳå£«

-0.71

ĻĤ

-0.70

spin

-0.68

deck

-0.65

friends

-0.65

pages

-0.64

shows

-0.64

worthy

-0.63

names

-0.63

POSITIVE LOGITS

ini

1.02

zzle

1.00

zzo

0.94

emi

0.94

etta

0.90

zzi

0.90

opsis

0.88

aga

0.88

ya

0.87

qi

0.86

Activations Density 0.010%

names, specifically last names

occurrences of the suffix 'ini' used in various contexts

No Comments

No Known Activations