INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

sequences of repeated tokens or placeholders

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.67

 choke

-0.66

 pockets

-0.61

 installation

-0.61

 transfers

-0.60

Ö¼

-0.60

 installations

-0.59

 Malfoy

-0.59

 folder

-0.59

 redemption

-0.59

POSITIVE LOGITS

Daily

1.13

aretz

1.12

News

1.11

Politics

1.09

NBC

1.03

 Politico

0.99

Journal

0.96

 Daily

0.96

daily

0.95

itbart

0.94

Activations Density 0.259%

proper nouns

sequences of repeated tokens or placeholders

No Comments

No Known Activations