INDEX

Explanations

instructions or steps starting with the word "To"

oai_token-act-pair · gpt-3.5-turbo

the phrase "To" at the beginning of sentences

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

soDeliveryDate

-0.80

 taps

-0.69

 looms

-0.63

 diapers

-0.63

hops

-0.62

itates

-0.62

 shaping

-0.61

dro

-0.60

 needed

-0.60

 nets

-0.59

POSITIVE LOGITS

ilet

1.31

asty

1.02

ffee

0.99

 summarize

0.99

pping

0.91

 reiterate

0.90

 clarify

0.90

 celebrate

0.89

ilers

0.89

pped

0.85

Activations Density 0.262%

instructions or steps starting with the word "To"

the phrase "To" at the beginning of sentences

No Comments

No Known Activations