INDEX

Explanations

words related to opinions, evaluations, and emotions towards various entities and topics

oai_token-act-pair · gpt-3.5-turbo

expressions of public opinion and trust in institutions

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Aim

-0.92

<?

-0.71

catentry

-0.70

Mesh

-0.70

wu

-0.69

alion

-0.67

enf

-0.66

yne

-0.65

pex

-0.65

çĶ°

-0.64

POSITIVE LOGITS

 their

1.03

 themselves

1.00

 THEIR

0.91

 politicians

0.90

their

0.86

 government

0.78

 Barack

0.78

government

0.78

 electing

0.78

 democracy

0.77

Activations Density 0.564%

words related to opinions, evaluations, and emotions towards various entities and topics

expressions of public opinion and trust in institutions

No Comments

No Known Activations