INDEX

Explanations

the word "Cohen"

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Cohen."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 reluct

-0.78

rha

-0.68

atical

-0.67

oppable

-0.67

ICAN

-0.65

 lockdown

-0.64

retched

-0.64

Results

-0.64

stood

-0.63

tub

-0.61

POSITIVE LOGITS

agher

0.98

²¾

0.90

bach

0.76

lette

0.72

¬¼

0.71

zon

0.70

wald

0.70

 Strauss

0.68

hiba

0.68

ombat

0.67

Activations Density 0.022%

the word "Cohen"

mentions of the name "Cohen."

No Comments

No Known Activations