INDEX

Explanations

words related to intentions or purposes

oai_token-act-pair · gpt-3.5-turbo

phrases that describe intended purposes or functions of various actions or items

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

 Solitaire

-0.81

yards

-0.79

tex

-0.68

las

-0.63

buck

-0.62

buster

-0.62

acks

-0.61

oner

-0.61

ranked

-0.61

browser

-0.60

POSITIVE LOGITS

 Parenthood

0.93

 intended

0.78

fully

0.76

 beneficiaries

0.74

atile

0.72

 recipients

0.70

ivil

0.70

 guiActiveUn

0.68

 beneficiary

0.66

lessly

0.66

Activations Density 0.033%

words related to intentions or purposes

phrases that describe intended purposes or functions of various actions or items

No Comments

No Known Activations