INDEX

Explanations

the specific word "Kremlin"

oai_token-act-pair · gpt-4-turbo

mentions of the word "Kremlin."

oai_token-act-pair · gpt-3.5-turbo

references to the Kremlin

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Thom

-0.78

ients

-0.75

 Kear

-0.73

matter

-0.72

ORGE

-0.70

 Scotch

-0.70

iencies

-0.70

 Subaru

-0.69

âĢ¢âĢ¢âĢ¢âĢ¢

-0.67

IRD

-0.67

POSITIVE LOGITS

 Kremlin

0.97

 rall

0.95

 dictatorship

0.88

achev

0.87

kaya

0.83

 dictator

0.81

istani

0.80

 disinformation

0.80

 affili

0.80

 Jinping

0.78

Activations Density 0.005%

the specific word "Kremlin"

mentions of the word "Kremlin."

references to the Kremlin

No Comments

No Known Activations