INDEX

Explanations

words indicating things or places named after individuals

oai_token-act-pair · gpt-3.5-turbo

instances of the word "named."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

yip

-0.90

tein

-0.67

veyard

-0.67

erers

-0.66

bs

-0.65

irez

-0.65

stead

-0.65

ijk

-0.64

plane

-0.64

psey

-0.62

POSITIVE LOGITS

 Cosponsors

0.81

paces

0.74

daq

0.72

 Tick

0.69

 plaintiff

0.66

lime

0.65

DAQ

0.65

 after

0.64

 Ones

0.62

 Priority

0.61

Activations Density 0.048%

words indicating things or places named after individuals

instances of the word "named."

No Comments

No Known Activations