INDEX

Explanations

instances of the word "Linden."

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Lind."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

OPLE

-0.82

CLASSIFIED

-0.78

Ø©

-0.74

Magikarp

-0.69

ABLE

-0.66

FTA

-0.64

 srfAttach

-0.62

 eleph

-0.60

IBLE

-0.60

voc

-0.59

POSITIVE LOGITS

gren

1.20

strom

1.10

isf

1.09

enstein

1.02

seys

1.01

say

0.96

holm

0.94

ahl

0.94

quist

0.94

lasses

0.93

Activations Density 0.013%

instances of the word "Linden."

mentions of the name "Lind."

No Comments

No Known Activations