INDEX

Explanations

instances where something is expected to happen

oai_token-act-pair · gpt-3.5-turbo

phrases indicating predictions or anticipations about future events

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

Interstitial

-0.76

backer

-0.74

 Disease

-0.70

cit

-0.66

ses

-0.64

reen

-0.63

tex

-0.63

tha

-0.62

ventions

-0.61

tein

-0.59

POSITIVE LOGITS

icipated

0.72

 applause

0.68

LY

0.66

eln

0.66

 newsp

0.65

 divest

0.65

 plur

0.64

 unanim

0.62

 depreciation

0.62

heny

0.61

Activations Density 0.025%

instances where something is expected to happen

phrases indicating predictions or anticipations about future events

No Comments

No Known Activations