INDEX

Explanations

names of individuals, particularly with a focus on "Rin" and "Nico"

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names related to individuals or entities

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

berra

-0.85

acular

-0.68

een

-0.68

mort

-0.68

ritional

-0.63

eneg

-0.63

ers

-0.63

yer

-0.62

rait

-0.62

stroke

-0.62

POSITIVE LOGITS

Â·Â·

0.83

itarian

0.83

Rin

0.81

Ig

0.81

za

0.69

Ã¡s

0.68

 Roosevelt

0.68

zl

0.67

ASED

0.67

zman

0.66

Activations Density 0.046%

names of individuals, particularly with a focus on "Rin" and "Nico"

proper nouns, particularly names related to individuals or entities

No Comments

No Known Activations