INDEX

Explanations

references to specific names or titles

oai_token-act-pair · gpt-3.5-turbo

references to specific entities, primarily names associated with media and veterinary contexts

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

ogly

-0.69

earance

-0.69

quartered

-0.69

eworld

-0.67

brook

-0.66

thritis

-0.66

athering

-0.64

athered

-0.64

oug

-0.63

gotten

-0.63

POSITIVE LOGITS

 Diesel

0.76

ideos

0.74

III

0.74

iolet

0.74

 Machina

0.73

icious

0.72

alkyrie

0.72

 Yanukovych

0.71

rine

0.71

istors

0.70

Activations Density 0.064%

references to specific names or titles

references to specific entities, primarily names associated with media and veterinary contexts

No Comments

No Known Activations