INDEX

Explanations

mentions of political figures and elections

oai_token-act-pair · gpt-3.5-turbo

structural elements of news articles or stories

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Stage

-0.18

 tremend

-0.16

Ye

-0.16

osate

-0.15

Py

-0.15

dyn

-0.15

 Race

-0.15

 Recall

-0.15

 noble

-0.15

Dyn

-0.14

POSITIVE LOGITS

ONSORED

0.22

 taxp

0.18

Reilly

0.18

soDeliveryDate

0.18

 outwe

0.17

ãĥ¼ãĥ³

0.17

uable

0.17

annie

0.17

 understatement

0.17

Activations Density 12.671%

mentions of political figures and elections

structural elements of news articles or stories

No Comments

No Known Activations