INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names of characters, brands, or cultural references

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

assetsadobe

-0.67

 unequiv

-0.63

HIP

-0.60

 uncons

-0.59

 Medicare

-0.59

 general

-0.54

 emails

-0.54

å¼

-0.52

Mb

-0.52

[];

-0.52

POSITIVE LOGITS

eware

0.83

astery

0.77

iken

0.77

utenberg

0.76

icycle

0.76

ipers

0.75

keyes

0.74

ilian

0.74

roller

0.73

iverpool

0.73

Activations Density 0.348%

proper nouns

proper nouns, particularly names of characters, brands, or cultural references

No Comments

No Known Activations