INDEX

Explanations

phrases describing intentions or plans involving specific actions

oai_token-act-pair · gpt-3.5-turbo

phrases that involve conditions, consequences, or actions related to policies and governance

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

ãĤ¦ãĤ¹

-0.69

quished

-0.68

urnal

-0.66

 Selected

-0.63

DX

-0.61

uber

-0.61

ĻĤ

-0.60

raved

-0.60

awed

-0.60

itarian

-0.59

POSITIVE LOGITS

 thereby

1.21

 eliminate

1.16

 lest

1.15

 preferably

1.13

 abolish

1.11

 stabilize

1.11

 dismantle

1.08

 consolidate

1.08

 establish

1.04

 weaken

1.04

Activations Density 0.319%

phrases describing intentions or plans involving specific actions

phrases that involve conditions, consequences, or actions related to policies and governance

No Comments

No Known Activations