INDEX

Explanations

numerical values that are markers or identifiers

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter "n" and related entities or terms with high frequency

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

NESS

-0.72

eve

-0.64

IFE

-0.61

ARDIS

-0.60

STATE

-0.58

 curing

-0.57

fet

-0.56

 snowball

-0.56

 cure

-0.56

 stewards

-0.55

POSITIVE LOGITS

esides

1.10

0.96

imately

0.92

umably

0.91

arently

0.91

ickets

0.90

idav

0.90

onymous

0.90

)(

0.88

ocated

0.88

Activations Density 0.123%

numerical values that are markers or identifiers

occurrences of the letter "n" and related entities or terms with high frequency

No Comments

No Known Activations