INDEX

Explanations

words related to emotions, beliefs, and attitudes towards authority or government

oai_token-act-pair · gpt-3.5-turbo

concepts related to trust, understanding, and emotional states in social and political contexts

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

aster

-0.85

ãĤ´ãĥ³

-0.77

dule

-0.73

 ç¥ŀ

-0.66

uddin

-0.66

poon

-0.66

 scissors

-0.64

icles

-0.64

fixes

-0.62

icle

-0.62

POSITIVE LOGITS

 among

1.21

 amongst

1.18

 perv

1.05

among

1.04

 prevalent

0.96

 expressed

0.92

 perme

0.85

 emanating

0.85

 evident

0.85

 prevailed

0.84

Activations Density 0.262%

words related to emotions, beliefs, and attitudes towards authority or government

concepts related to trust, understanding, and emotional states in social and political contexts

No Comments

No Known Activations