INDEX

Explanations

words related to accusations or charges, specifically when they are serious

oai_token-act-pair · gpt-3.5-turbo

the term "allegations" and its related context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

patch

-0.73

skill

-0.71

oppy

-0.68

tz

-0.68

omez

-0.68

ga

-0.67

 busiest

-0.67

keys

-0.65

kin

-0.65

cise

-0.63

POSITIVE LOGITS

 allegations

1.14

 accusations

1.00

 alleging

0.96

 allegation

0.95

 accusation

0.91

 alleges

0.89

 accusing

0.87

 accus

0.78

 levied

0.77

 leveled

0.75

Activations Density 0.031%

words related to accusations or charges, specifically when they are serious

the term "allegations" and its related context

No Comments

No Known Activations