INDEX

Explanations

words related to locations or events

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "in"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

TING

-0.61

bryce

-0.61

 narrator

-0.59

 Kislyak

-0.59

alist

-0.59

assetsadobe

-0.58

lun

-0.55

 average

-0.55

ichick

-0.55

 mentally

-0.54

POSITIVE LOGITS

forced

1.11

hardt

1.05

hart

1.04

strument

1.00

struct

0.94

ews

0.94

forcement

0.94

vent

0.91

forcing

0.91

heit

0.90

Activations Density 0.035%

words related to locations or events

occurrences of the word "in"

No Comments

No Known Activations