INDEX

Explanations

repeated mentions of specific personal names

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

names of individuals or characters

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

intent

-0.74

pmwiki

-0.74

ERAL

-0.66

 Tyrann

-0.64

gio

-0.64

 Aliens

-0.63

mbol

-0.63

egal

-0.61

ãĥķãĤ©

-0.61

tenance

-0.61

POSITIVE LOGITS

 Owen

1.09

oken

0.89

Ree

0.85

shire

0.81

terness

0.79

oven

0.78

bridge

0.76

kefeller

0.76

 Farrell

0.75

 Owens

0.75

Activations Density 0.011%

repeated mentions of specific personal names

names of individuals or characters

No Comments

No Known Activations