INDEX

Explanations

numbers related to quantities or measurements

oai_token-act-pair · gpt-3.5-turbo

verbs indicating actions or events occurring in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

thur

-0.78

hack

-0.68

thin

-0.67

nown

-0.65

ixel

-0.64

idium

-0.64

Phys

-0.64

Enlarge

-0.63

jee

-0.63

wr

-0.62

POSITIVE LOGITS

UCT

0.68

ÃĥÃĤ

0.68

 subur

0.68

 satell

0.64

 referen

0.64

escription

0.63

 separately

0.62

nesday

0.61

 pione

0.61

 tradem

0.61

Activations Density 0.474%

numbers related to quantities or measurements

verbs indicating actions or events occurring in various contexts

No Comments

No Known Activations