INDEX

Explanations

phrases indicating preference for a simpler or more direct approach

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "just."

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

 challeng

-0.72

 Palestin

-0.67

bestos

-0.62

 PLUS

-0.62

antage

-0.62

 protector

-0.62

 Remastered

-0.61

ikuman

-0.61

 adversary

-0.61

luster

-0.60

POSITIVE LOGITS

ifiable

1.15

ifications

1.06

if

0.98

ifi

0.95

ified

0.82

ify

0.82

itia

0.80

IFIC

0.79

ifiers

0.78

icia

0.78

Activations Density 0.095%

phrases indicating preference for a simpler or more direct approach

occurrences of the word "just."

No Comments

No Known Activations