INDEX

Explanations

phrases referring to specific actions or events

oai_token-act-pair · gpt-3.5-turbo

indicative phrases suggesting actions or decisions by groups or individuals

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

.","

-0.55

zag

-0.52

ige

-0.51

=-=-=-=-=-=-=-=-

-0.49

 UCHIJ

-0.49

apan

-0.49

aer

-0.48

ixir

-0.48

iership

-0.48

cients

-0.47

POSITIVE LOGITS

 latter

0.64

 irony

0.63

atto

0.60

 scathing

0.55

 damning

0.54

 NOTICE

0.53

arently

0.52

 excerpts

0.52

 furthermore

0.51

 ironic

0.51

Activations Density 1.661%

phrases referring to specific actions or events

indicative phrases suggesting actions or decisions by groups or individuals

No Comments

No Known Activations