INDEX

Explanations

words related to negative or critical evaluations of events, actions, or people

oai_token-act-pair · gpt-3.5-turbo

negative prefixes or terms related to mistakes and mismanagement

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 crates

-0.81

Ô

-0.78

 kinds

-0.71

theless

-0.71

 helicopters

-0.70

 registrations

-0.70

 clients

-0.70

 equivalents

-0.69

 defects

-0.69

 requests

-0.69

POSITIVE LOGITS

anza

1.06

agonist

1.03

insula

1.01

omination

1.00

stros

0.98

ologue

0.97

hander

0.96

undrum

0.96

keleton

0.95

agogue

0.95

Activations Density 0.548%

words related to negative or critical evaluations of events, actions, or people

negative prefixes or terms related to mistakes and mismanagement

No Comments

No Known Activations