INDEX

Explanations

references to a specific person named Warren

oai_token-act-pair · gpt-3.5-turbo

occurrences of the name "Warren."

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

dayName

-0.88

ãĤ´ãĥ³

-0.82

netflix

-0.79

pmwiki

-0.79

odic

-0.73

unal

-0.73

ãĤº

-0.70

asus

-0.68

cffff

-0.67

mble

-0.67

POSITIVE LOGITS

 Buffett

1.15

sburg

1.10

 Warren

1.01

 Farrell

0.93

ers

0.88

rade

0.88

ston

0.87

flower

0.78

 Sapp

0.78

 Davidson

0.77

Activations Density 0.013%

references to a specific person named Warren

occurrences of the name "Warren."

No Comments

No Known Activations