INDEX

Explanations

phrases related to intellectual or philosophical discussions and concepts

oai_token-act-pair · gpt-3.5-turbo

complex ideas and concepts related to critical analysis and evaluation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

VL

-0.65

 notor

-0.62

swick

-0.60

$.

-0.59

essage

-0.58

kef

-0.58

ãĥ¯ãĥ³

-0.57

flix

-0.57

Interstitial

-0.56

wikipedia

-0.56

POSITIVE LOGITS

 awaits

0.69

Â¶

0.68

 âĵĺ

0.59

"?

0.59

 prompt

0.56

 leaps

0.55

 aside

0.54

 Posted

0.54

huh

0.53

 Dragonbound

0.52

Activations Density 0.463%

phrases related to intellectual or philosophical discussions and concepts

complex ideas and concepts related to critical analysis and evaluation

No Comments

No Known Activations