INDEX

Explanations

persons or groups who are threatened or face consequences if they do not comply with specific actions

oai_token-act-pair · gpt-3.5-turbo

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

IFrame

Link

Negative Logits

>>>>>>>>

-0.77

 Selection

-0.71

 optimum

-0.69

 Lifetime

-0.67

 Surviv

-0.66

 Mood

-0.64

Psych

-0.63

aukee

-0.63

iple

-0.63

inct

-0.63

POSITIVE LOGITS

 dared

1.25

 disob

1.23

 disagree

1.18

 disple

1.12

 disagreed

1.11

 dare

1.08

 transgress

1.05

 refused

1.05

 defied

1.05

 resisted

1.01

Activations Density 0.294%

persons or groups who are threatened or face consequences if they do not comply with specific actions

No Known Activations