INDEX

Explanations

texts related to actions or decisions

oai_token-act-pair · gpt-3.5-turbo

the presence of colons

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 troubles

-0.74

undai

-0.72

yan

-0.72

roit

-0.69

entle

-0.64

apan

-0.62

leased

-0.61

IPM

-0.60

 tremend

-0.60

 overl

-0.59

POSITIVE LOGITS

 Provided

1.12

 namely

1.03

 Journals

0.78

 http

0.76

 https

0.76

↵Âł

0.68

 Whereas

0.67

 "âĢ¦

0.65

 Firstly

0.64

until

0.64

Activations Density 0.120%

texts related to actions or decisions

the presence of colons

No Comments

No Known Activations