INDEX

Explanations

topics related to personal privacy and harm

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.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

 Cerberus

-0.74

SPONSORED

-0.67

ndum

-0.67

else

-0.66

Reply

-0.66

"))

-0.66

 Â¯

-0.64

reverse

-0.60

 Nare

-0.59

RESULTS

-0.58

POSITIVE LOGITS

*,

0.64

 familiarity

0.60

vs

0.60

 stiffness

0.59

Â®,

0.57

 versus

0.57

 charisma

0.57

 clutch

0.57

 spouse

0.57

 scarcity

0.57

Activations Density 0.513%

topics related to personal privacy and harm

No Comments

No Known Activations