INDEX

Explanations

mentions of politicians

oai_token-act-pair · gpt-3.5-turbo

terms related to political discussions and commentary

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

uran

-0.72

 Warrant

-0.71

LEASE

-0.70

upon

-0.70

RED

-0.69

 pity

-0.68

urous

-0.68

VIEW

-0.68

unction

-0.67

orney

-0.66

POSITIVE LOGITS

icians

1.07

ifact

1.03

 Polit

1.01

ician

0.95

icial

0.94

ically

0.88

 correctness

0.85

eness

0.84

 Pengu

0.77

 paran

0.76

Activations Density 0.009%

mentions of politicians

terms related to political discussions and commentary

No Comments

No Known Activations