INDEX

Explanations

phrases including the word "leader."

oai_token-act-pair · gpt-3.5-turbo

occurrences of the suffix "er" 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 @ 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

ĸļ

-0.96

acebook

-0.71

chwitz

-0.69

eenth

-0.69

raltar

-0.68

atform

-0.65

luaj

-0.62

urities

-0.62

ertodd

-0.61

uncture

-0.60

POSITIVE LOGITS

jee

1.01

lein

0.95

idge

0.88

ger

0.86

adish

0.84

lich

0.82

rors

0.82

aton

0.81

baum

0.79

asures

0.79

Activations Density 0.051%

phrases including the word "leader."

occurrences of the suffix "er" in various contexts

No Comments

No Known Activations