INDEX

Explanations

words related to medical, geographical or personal names

oai_token-act-pair · gpt-3.5-turbo

the word "one" 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 @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 actionGroup

-0.73

 precip

-0.65

hips

-0.64

 resil

-0.63

achusetts

-0.62

NRS

-0.62

lished

-0.61

 brig

-0.61

æĸ¹

-0.59

 prof

-0.58

POSITIVE LOGITS

xus

0.99

gger

0.95

xit

0.93

ones

0.90

one

0.90

hoe

0.85

boarding

0.82

ople

0.78

ciating

0.78

address

0.76

Activations Density 0.012%

words related to medical, geographical or personal names

the word "one" in various contexts

No Comments

No Known Activations