INDEX

Explanations

specific named entities related to politics and medical terms

oai_token-act-pair · gpt-3.5-turbo

references to identities and geographic locations associated with societal structures

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

eering

-0.74

aku

-0.67

stakes

-0.65

ing

-0.65

atility

-0.64

Kov

-0.62

 Antar

-0.62

dogs

-0.62

enthal

-0.62

 agre

-0.61

POSITIVE LOGITS

itate

0.89

illin

0.81

forth

0.77

alyst

0.76

sei

0.76

itation

0.75

itated

0.75

onymous

0.74

herent

0.74

esville

0.73

Activations Density 0.051%

specific named entities related to politics and medical terms

references to identities and geographic locations associated with societal structures

No Comments

No Known Activations