INDEX

Explanations

numbers in the text

oai_token-act-pair · gpt-3.5-turbo

references to statistics or numerical totals

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

tera

-0.69

soType

-0.68

 privile

-0.63

era

-0.63

hend

-0.62

 Receiver

-0.62

riers

-0.60

Bomb

-0.59

 Branch

-0.58

 Quote

-0.58

POSITIVE LOGITS

 sorts

0.90

 eighty

0.73

oscopic

0.69

 pounds

0.69

 course

0.68

0.67

abama

0.66

 sixty

0.66

 seventy

0.66

 eight

0.66

Activations Density 0.070%

numbers in the text

references to statistics or numerical totals

No Comments

No Known Activations