INDEX

Explanations

phrases related to simple versus complicated processes or concepts

oai_token-act-pair · gpt-3.5-turbo

complex and ambiguous phrases or structures

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

Not in Any Lists

No Comments

Negative Logits

chwitz

-0.78

itton

-0.77

isition

-0.74

ersion

-0.74

AMD

-0.73

ilers

-0.73

akery

-0.73

ensable

-0.72

auld

-0.72

orsi

-0.71

POSITIVE LOGITS

Sno

0.71

tha

0.70

uph

0.69

 Lucy

0.66

ta

0.66

mah

0.65

 thee

0.65

 Into

0.65

pu

0.63

lil

0.63

Activations Density 0.449%

phrases related to simple versus complicated processes or concepts

complex and ambiguous phrases or structures

No Comments

No Known Activations