INDEX

Explanations

terms related to liberal ideologies

oai_token-act-pair · gpt-3.5-turbo

references to liberal ideology and associated concepts

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

iru

-0.76

pty

-0.68

atoon

-0.67

BLE

-0.66

breaker

-0.66

angan

-0.66

 Danger

-0.65

 Blazing

-0.65

atel

-0.64

abba

-0.63

POSITIVE LOGITS

ization

1.27

izing

1.17

izers

1.13

 arts

1.08

ized

1.07

izations

1.04

izes

1.01

isation

1.00

izer

0.99

ize

0.93

Activations Density 0.024%

terms related to liberal ideologies

references to liberal ideology and associated concepts

No Comments

No Known Activations