INDEX

Explanations

mentions of specific apps or software platforms

oai_token-act-pair · gpt-3.5-turbo

references to digital applications and notable figures

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

 charism

-0.79

 stock

-0.67

CVE

-0.67

 equity

-0.65

 stake

-0.64

yip

-0.64

 coined

-0.63

 understatement

-0.63

 Helpful

-0.61

entious

-0.61

POSITIVE LOGITS

atern

0.86

SIGN

0.75

numbered

0.75

blance

0.69

agonists

0.68

letters

0.67

Pass

0.66

grad

0.66

mson

0.66

adra

0.66

Activations Density 0.000%

mentions of specific apps or software platforms

references to digital applications and notable figures

No Comments

No Known Activations