INDEX

Explanations

a specific uppercase letter 'L' with different variations and contexts

oai_token-act-pair · gpt-3.5-turbo

references to a specific individual or entity labeled "L."

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

ãĥ¼ãĥĨãĤ£

-0.81

rolet

-0.75

smanship

-0.72

hower

-0.70

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.69

 Subtle

-0.69

 Primal

-0.67

gerald

-0.66

ãĤ¡

-0.65

ĳå£«

-0.65

POSITIVE LOGITS

ocker

1.23

ifestyle

1.17

ibrarian

1.16

adders

1.13

aund

1.11

ipp

1.09

aptop

1.09

ayers

1.08

umber

1.07

arger

1.07

Activations Density 0.043%

a specific uppercase letter 'L' with different variations and contexts

references to a specific individual or entity labeled "L."

No Comments

No Known Activations