INDEX

Explanations

words expressing negative emotions or outcomes

oai_token-act-pair · gpt-3.5-turbo

expressions of regret or misfortune

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

arnaev

-0.78

addons

-0.69

rouse

-0.67

ĸļ

-0.67

rounder

-0.67

aver

-0.66

aeda

-0.64

kefeller

-0.64

afort

-0.64

arij

-0.64

POSITIVE LOGITS

0.91

 enough

0.90

,...

0.80

for

0.74

 though

0.74

 neither

0.73

 however

0.70

 alas

0.69

 there

0.68

 none

0.67

Activations Density 0.058%

words expressing negative emotions or outcomes

expressions of regret or misfortune

No Comments

No Known Activations