INDEX

Explanations

proper nouns or names

oai_token-act-pair · gpt-3.5-turbo

mentions of specific names, particularly the name "Abe" 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 @ 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

phis

-0.96

imates

-0.87

neapolis

-0.87

ivities

-0.82

ileaks

-0.82

angular

-0.81

prus

-0.80

matic

-0.80

ophical

-0.80

imedia

-0.80

POSITIVE LOGITS

zz

0.84

legates

0.81

legate

0.79

FORE

0.76

zza

0.76

deen

0.71

zzi

0.71

Ã§a

0.69

gger

0.69

cki

0.69

Activations Density 0.037%

proper nouns or names

mentions of specific names, particularly the name "Abe" in various contexts

No Comments

No Known Activations