INDEX

Explanations

phrases related to origins or causes of events

oai_token-act-pair · gpt-3.5-turbo

phrases indicating emergence or departure from a state

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

ļéĨĴ

-0.87

cious

-0.76

Export

-0.72

anooga

-0.70

EY

-0.69

ancial

-0.65

ingham

-0.65

uyomi

-0.65

Tips

-0.63

 Illustrated

-0.63

POSITIVE LOGITS

wards

0.86

fitted

0.82

stretched

0.77

doors

0.75

worn

0.74

flows

0.73

casts

0.70

doing

0.70

stri

0.70

breaks

0.68

Activations Density 0.055%

phrases related to origins or causes of events

phrases indicating emergence or departure from a state

No Comments

No Known Activations