INDEX

Explanations

information about negotiations or discussions happening between entities

oai_token-act-pair · gpt-3.5-turbo

the preposition "in."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 alloc

-0.68

uces

-0.68

 ethic

-0.61

 intimid

-0.59

irl

-0.59

Voice

-0.58

NOW

-0.57

 Doodle

-0.56

CRIPTION

-0.56

 Sorceress

-0.54

POSITIVE LOGITS

escap

1.18

ching

1.16

strument

1.06

humane

1.05

operative

1.04

patient

1.03

bred

1.00

 jeopardy

0.99

ked

0.99

animate

0.98

Activations Density 0.158%

information about negotiations or discussions happening between entities

the preposition "in."

No Comments

No Known Activations