INDEX

Explanations

phrases related to confessing or admitting something

oai_token-act-pair · gpt-3.5-turbo

the phrase "to" indicating actions or admissions made by individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

estyles

-0.75

served

-0.74

blooded

-0.67

ahead

-0.67

interstitial

-0.67

edged

-0.66

nai

-0.65

 è£ıè¦ļéĨĴ

-0.65

clips

-0.63

 appealed

-0.63

POSITIVE LOGITS

us

0.91

ilet

0.86

me

0.85

psy

0.79

ggles

0.73

Ĭ±

0.69

pless

0.66

asted

0.66

 congress

0.64

orians

0.61

Activations Density 0.099%

phrases related to confessing or admitting something

the phrase "to" indicating actions or admissions made by individuals

No Comments

No Known Activations