INDEX

Explanations

phrases related to requests or commands

oai_token-act-pair · gpt-3.5-turbo

phrases that involve directives or calls to action

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

 sqor

-0.70

 srfAttach

-0.69

ivan

-0.68

 Crimson

-0.66

PsyNetMessage

-0.64

luck

-0.64

 partName

-0.64

 Genius

-0.62

 Wraith

-0.62

 Dragonbound

-0.62

POSITIVE LOGITS

 recons

1.19

 replaced

1.14

 treated

1.11

 abolished

1.09

 reinstated

1.06

 avoided

1.04

 considered

1.03

 scrapped

1.01

 restored

1.01

 preserved

1.01

Activations Density 0.168%

phrases related to requests or commands

phrases that involve directives or calls to action

No Comments

No Known Activations