INDEX

Explanations

names of Russian and Ukrainian individuals

oai_token-act-pair · gpt-3.5-turbo

names of individuals, particularly those of Russian origin

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

nels

-0.82

ãĥ¼ãĥĨ

-0.76

apple

-0.74

 Fiesta

-0.73

plain

-0.73

finding

-0.72

isons

-0.72

roads

-0.70

neys

-0.68

cess

-0.68

POSITIVE LOGITS

 Mikhail

1.10

 Lavrov

1.09

 Dmitry

1.02

 Sergei

0.99

 Putin

0.98

ovych

0.96

 Sergey

0.95

 Gork

0.95

 Dmit

0.93

Ily

0.89

Activations Density 0.029%

names of Russian and Ukrainian individuals

names of individuals, particularly those of Russian origin

No Comments

No Known Activations