INDEX

Explanations

references to entities named 'No' with accompanying numbers

oai_token-act-pair · gpt-3.5-turbo

references to a specific entity labeled with "No" followed by a number

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

staking

-0.76

theless

-0.72

 priceless

-0.71

 insecure

-0.64

 muted

-0.64

 unavailable

-0.62

 outwe

-0.60

romeda

-0.59

 tong

-0.58

FML

-0.58

POSITIVE LOGITS

emi

1.10

zzle

1.07

isy

1.05

onday

1.01

isec

0.98

vel

0.94

Ã«

0.91

xious

0.90

omi

0.87

Va

0.87

Activations Density 0.033%

references to entities named 'No' with accompanying numbers

references to a specific entity labeled with "No" followed by a number

No Comments

No Known Activations