INDEX

Explanations

references to things being released or put out into the world

oai_token-act-pair · gpt-3.5-turbo

phrases that include the word "out."

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

cious

-0.73

avorite

-0.70

iosity

-0.68

jriwal

-0.68

interstitial

-0.60

VK

-0.59

xit

-0.58

ugh

-0.57

 turnover

-0.57

antry

-0.56

POSITIVE LOGITS

fitted

1.01

lier

0.92

flows

0.92

smart

0.85

stretched

0.84

posts

0.80

lived

0.80

lander

0.79

doors

0.75

fitting

0.75

Activations Density 0.077%

references to things being released or put out into the world

phrases that include the word "out."

No Comments

No Known Activations