INDEX

Explanations

phrases related to news stories and reports

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

references to audience engagement or interaction

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

leeve

-0.60

 relative

-0.60

iliated

-0.59

¿½

-0.57

 Citation

-0.56

VC

-0.56

Fax

-0.55

SIL

-0.55

 comparison

-0.55

 buck

-0.55

POSITIVE LOGITS

abad

0.80

sbm

0.78

ilst

0.74

combe

0.70

tw

0.68

rise

0.67

tf

0.67

abases

0.66

 externalToEVAOnly

0.64

Îµ

0.64

Activations Density 0.156%

phrases related to news stories and reports

references to audience engagement or interaction

No Comments

No Known Activations