INDEX

Explanations

names, likely of people

oai_token-act-pair · gpt-3.5-turbo

the mention of specific names or identities

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

nces

-0.80

ILY

-0.79

minist

-0.78

rance

-0.76

scl

-0.75

gged

-0.75

cale

-0.74

gently

-0.74

INESS

-0.74

backer

-0.74

POSITIVE LOGITS

atalie

0.80

 Devi

0.78

 Loren

0.76

eus

0.74

 Brus

0.73

itri

0.67

Nik

0.66

udeau

0.65

 Port

0.65

 Natalie

0.64

Activations Density 0.043%

names, likely of people

the mention of specific names or identities

No Comments

No Known Activations