INDEX

Explanations

proper nouns, specifically related to political figures

oai_token-act-pair · gpt-3.5-turbo

references to specific situations or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

EEE

-0.69

cffff

-0.68

heon

-0.67

pherd

-0.65

soType

-0.65

axter

-0.64

eln

-0.63

wcsstore

-0.63

ð

-0.62

à¹

-0.62

POSITIVE LOGITS

istan

0.82

aires

0.77

rift

0.71

tab

0.68

igmatic

0.67

off

0.66

rama

0.66

 Instit

0.66

zan

0.64

sag

0.63

Activations Density 0.000%

proper nouns, specifically related to political figures

references to specific situations or events

No Comments

No Known Activations