INDEX

Explanations

keywords signalling the start of a section or list in a document

oai_token-act-pair · gpt-3.5-turbo

sections of content labeled with the word "Below," indicating subsections or lists within a document

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

natureconservancy

-0.75

Roh

-0.66

ifiable

-0.65

aceutical

-0.65

ãĥı

-0.65

ãĤ£

-0.60

imm

-0.57

ãĤ¹ãĥĪ

-0.57

hya

-0.57

aternity

-0.55

POSITIVE LOGITS

neath

0.89

 Thumbnails

0.84

ground

0.81

İĭ

0.75

rites

0.75

henko

0.71

deck

0.68

peak

0.68

fter

0.68

 veter

0.68

Activations Density 0.024%

keywords signalling the start of a section or list in a document

sections of content labeled with the word "Below," indicating subsections or lists within a document

No Comments

No Known Activations