INDEX

Explanations

phrases related to holding law enforcement accountable

oai_token-act-pair · gpt-3.5-turbo

negative sentiment or criticism

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

DD

-0.78

BB

-0.74

HT

-0.73

 Feder

-0.73

 Simon

-0.72

 Fitz

-0.72

 Nish

-0.71

BBC

-0.70

Cow

-0.70

Bun

-0.70

POSITIVE LOGITS

bodied

1.17

axis

1.16

family

1.15

direction

1.14

induced

1.13

deal

1.13

based

1.12

centered

1.10

inspired

1.10

talk

1.10

Activations Density 0.095%

phrases related to holding law enforcement accountable

negative sentiment or criticism

No Comments

No Known Activations