INDEX

Explanations

initials followed by a numerical value or word, with a hyphen sometimes included

oai_token-act-pair · gpt-3.5-turbo

the letter "T" 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 @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 vacancy

-0.69

 exha

-0.65

 passing

-0.62

 Witch

-0.59

 referrals

-0.59

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.57

 livest

-0.57

 rounding

-0.57

 minded

-0.57

 Atmosp

-0.56

POSITIVE LOGITS

ARGET

1.35

reatment

1.31

ottenham

1.30

ruly

1.29

ravis

1.28

oward

1.28

itled

1.27

otally

1.26

utorial

1.26

ribune

1.26

Activations Density 0.040%

initials followed by a numerical value or word, with a hyphen sometimes included

the letter "T" in various contexts

No Comments

No Known Activations