INDEX

Explanations

names ending in 'ya'

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "ya" 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 @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

yright

-0.85

rodu

-0.79

OWER

-0.75

olation

-0.73

elong

-0.72

igree

-0.71

TAIN

-0.70

ochet

-0.70

matically

-0.70

opher

-0.67

POSITIVE LOGITS

 Pradesh

0.99

 Sabha

0.92

zza

0.88

kees

0.84

Yug

0.83

Nik

0.81

ichi

0.80

Suk

0.78

 nerv

0.78

ccess

0.78

Activations Density 0.012%

names ending in 'ya'

occurrences of the word "ya" in various contexts

No Comments

No Known Activations