INDEX

Explanations

references to body parts or physical locations on the body, specifically focusing on the back

oai_token-act-pair · gpt-3.5-turbo

the word "back" and its variations in various contexts

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

inational

-0.83

ISON

-0.69

aukee

-0.67

entric

-0.67

uary

-0.66

ciation

-0.65

Osc

-0.65

 Reloaded

-0.61

orpor

-0.60

uria

-0.60

POSITIVE LOGITS

side

1.22

stab

1.17

woods

1.14

packs

1.13

bones

1.10

GROUND

1.06

stories

1.06

seat

1.06

door

1.06

haul

1.04

Activations Density 0.049%

references to body parts or physical locations on the body, specifically focusing on the back

the word "back" and its variations in various contexts

No Comments

No Known Activations