INDEX

Explanations

words ending in "ol" or "ole"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "ol" 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

DATA

-0.64

æĿ

-0.63

âĸĳâĸĳ

-0.63

 Shards

-0.62

 Debor

-0.61

 Swanson

-0.60

 Citizen

-0.58

Glass

-0.57

Factor

-0.57

 Bigfoot

-0.56

POSITIVE LOGITS

ipop

1.25

ateral

1.19

onial

1.08

phe

1.07

ounge

1.06

ocation

1.01

okia

1.00

oad

0.99

anguage

0.97

anguages

0.96

Activations Density 0.043%

words ending in "ol" or "ole"

occurrences of the word "ol" in various contexts

No Comments

No Known Activations