INDEX

Explanations

words related to being beholden or subjected to certain circumstances or deals

oai_token-act-pair · gpt-3.5-turbo

the presence of the word "en" in various contexts

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

 Canaver

-0.79

enhagen

-0.76

itars

-0.73

mson

-0.71

orsche

-0.70

 tremend

-0.69

ONSORED

-0.68

atography

-0.68

ython

-0.68

 skelet

-0.67

POSITIVE LOGITS

berger

1.01

wald

0.93

feld

0.90

emies

0.89

heim

0.88

baum

0.88

heimer

0.86

bach

0.84

baugh

0.83

heit

0.81

Activations Density 0.027%

words related to being beholden or subjected to certain circumstances or deals

the presence of the word "en" in various contexts

No Comments

No Known Activations