INDEX

Explanations

the word "hope" with strong activation values

oai_token-act-pair · gpt-3.5-turbo

expressions conveying hope

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

cise

-0.81

versions

-0.72

Interstitial

-0.69

pes

-0.69

geant

-0.68

cers

-0.66

 masse

-0.64

ve

-0.63

å¥³

-0.63

VERTISEMENT

-0.63

POSITIVE LOGITS

lessly

1.06

fulness

0.93

lessness

0.90

eful

0.79

fully

0.77

bringer

0.76

ful

0.75

 awaited

0.74

 Solo

0.74

 hope

0.72

Activations Density 0.036%

the word "hope" with strong activation values

expressions conveying hope

No Comments

No Known Activations