INDEX

Explanations

phrases related to allegations or evidence towards a specific political figure

oai_token-act-pair · gpt-3.5-turbo

phrases related to political evidence and belief systems

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

 corrections

-0.66

IOC

-0.57

Äĩ

-0.57

vier

-0.57

eter

-0.55

 Correction

-0.54

theless

-0.54

ADS

-0.53

Ã©t

-0.52

 FranÃ§

-0.50

POSITIVE LOGITS

 Fract

0.57

åĤ

0.56

 Rosenstein

0.56

 Kavanaugh

0.55

 Meadows

0.55

ħ

0.55

Ĥ¬

0.55

Ń·

0.54

¬¼

0.54

amera

0.53

Activations Density 0.021%

phrases related to allegations or evidence towards a specific political figure

phrases related to political evidence and belief systems

No Comments

No Known Activations