INDEX

Explanations

phrases indicating a call to action or prompting an action to be taken

oai_token-act-pair · gpt-3.5-turbo

sequences of actions or responsibilities associated with regulation or advocacy

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

shire

-0.60

SourceFile

-0.59

tons

-0.58

orses

-0.57

inois

-0.57

lip

-0.56

skinned

-0.55

doc

-0.55

modified

-0.54

Letter

-0.54

POSITIVE LOGITS

wered

1.07

ggles

0.99

 facilitate

0.94

pless

0.94

adies

0.91

 reiterate

0.90

 promote

0.90

ffee

0.89

 conserve

0.89

 provide

0.89

Activations Density 0.089%

phrases indicating a call to action or prompting an action to be taken

sequences of actions or responsibilities associated with regulation or advocacy

No Comments

No Known Activations