INDEX

Explanations

string patterns and formatting related to coding languages

oai_token-act-pair · gpt-3.5-turbo

symbols and formatting markers commonly used in data expressions

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

IFrame

Link

Negative Logits

rolet

-0.71

unia

-0.70

 force

-0.70

ichick

-0.68

inator

-0.68

selage

-0.68

ahime

-0.68

seless

-0.68

ivating

-0.67

ivated

-0.66

POSITIVE LOGITS

 Shards

0.93

(*

0.80

(*

0.80

âĢł

0.75

.*

0.72

Constructed

0.72

TON

0.70

/-

0.68

bah

0.67

FIELD

0.66

Activations Density 0.018%

string patterns and formatting related to coding languages

symbols and formatting markers commonly used in data expressions

No Known Activations