INDEX

Explanations

proper nouns and numbers in text

oai_token-act-pair · gpt-3.5-turbo

references to personal memories or experiences

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

 Miliband

-0.60

 gard

-0.58

 Assange

-0.57

 Fail

-0.57

 Labour

-0.57

 aggress

-0.55

 Tsarnaev

-0.54

 scapego

-0.54

 Ukip

-0.53

 Petraeus

-0.52

POSITIVE LOGITS

zona

0.73

çĶ°

0.66

Interstitial

0.61

 respectively

0.61

pts

0.58

ukemia

0.56

Ü

0.56

itled

0.56

ãĥīãĥ©ãĤ´ãĥ³

0.56

ĸļ

0.54

Activations Density 2.776%

proper nouns and numbers in text

references to personal memories or experiences

No Comments

No Known Activations