INDEX

Explanations

names related to politics and healthcare

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals, particularly politicians

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

rained

-0.75

————————

-0.72

rex

-0.72

Meta

-0.70

 mast

-0.70

company

-0.67

Ø§ÙĦ

-0.66

ĸļ

-0.66

oleon

-0.64

emet

-0.62

POSITIVE LOGITS

veyard

0.84

ayson

0.84

 Grimes

0.82

iffin

0.81

banks

0.79

saf

0.77

atures

0.75

roots

0.74

veland

0.74

bly

0.73

Activations Density 0.018%

names related to politics and healthcare

references to specific individuals, particularly politicians

No Comments

No Known Activations