INDEX

Explanations

expressions involving multiplication or increase, often with the term "fold"

oai_token-act-pair · gpt-3.5-turbo

references to multiplicative concepts or repetitions

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

assetsadobe

-0.71

IRO

-0.69

FTWARE

-0.69

BAT

-0.69

nesota

-0.69

ulia

-0.68

igslist

-0.68

usalem

-0.67

Dialogue

-0.67

acea

-0.66

POSITIVE LOGITS

fold

1.20

ername

0.89

 fold

0.88

ers

0.85

edly

0.74

clip

0.73

cation

0.72

cause

0.71

theless

0.71

forward

0.70

Activations Density 0.012%

expressions involving multiplication or increase, often with the term "fold"

references to multiplicative concepts or repetitions

No Comments

No Known Activations