INDEX

Explanations

words related to car models and concepts

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "ro" 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 @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 EntityItem

-0.85

 Coulter

-0.83

ãĥ¼ãĥĨãĤ£

-0.75

CBI

-0.73

 Sons

-0.71

 Titanic

-0.65

 newsletters

-0.64

 eminent

-0.63

 Yankee

-0.62

 Catalyst

-0.62

POSITIVE LOGITS

spective

1.12

bably

1.12

active

1.05

perties

1.04

blems

1.01

tto

1.00

ving

0.98

bot

0.98

ggie

0.97

spection

0.97

Activations Density 0.016%

words related to car models and concepts

occurrences of the word "ro" in various contexts

No Comments

No Known Activations