INDEX

Explanations

expressions related to actions or intentions

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "just to" followed by various actions or clarifications

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

scar

-0.66

soever

-0.64

raising

-0.63

dominated

-0.61

significant

-0.60

ahime

-0.60

 Responsibility

-0.59

stri

-0.59

İĭ

-0.59

 Archdemon

-0.59

POSITIVE LOGITS

 amuse

1.04

 remind

1.00

 prove

0.97

 appease

0.96

 annoy

0.95

 piss

0.94

 satisfy

0.94

 reiterate

0.93

 keep

0.92

 reassure

0.92

Activations Density 0.078%

expressions related to actions or intentions

instances of the phrase "just to" followed by various actions or clarifications

No Comments

No Known Activations