INDEX

Explanations

information related to quantities and prices

oai_token-act-pair · gpt-3.5-turbo

numerical values and associated quantities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 ÂŃ

-0.66

 paramilitary

-0.63

eering

-0.61

 counterterrorism

-0.59

recy

-0.59

 solicit

-0.59

 totalitarian

-0.59

 postwar

-0.59

 instituted

-0.59

 solicitation

-0.58

POSITIVE LOGITS

mins

1.40

$.

1.34

yrs

1.34

+.

1.33

lbs

1.32

 mins

1.31

1.25

$,

1.24

1.22

$$

1.20

Activations Density 0.237%

information related to quantities and prices

numerical values and associated quantities

No Comments

No Known Activations