INDEX

Explanations

identification (ID) numbers

oai_token-act-pair · gpt-3.5-turbo

identifiers or codes, particularly related to organizational or location references

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

andise

-0.92

bilt

-0.81

 Seym

-0.74

 silence

-0.69

perty

-0.69

=-=-

-0.68

ussen

-0.65

 pandemonium

-0.62

¬¼

-0.62

 Sabb

-0.61

POSITIVE LOGITS

irect

1.22

DEN

1.10

iots

1.09

ocument

1.03

irection

1.03

GES

0.94

ouble

0.88

GE

0.88

IVERS

0.88

DER

0.87

Activations Density 0.010%

identification (ID) numbers

identifiers or codes, particularly related to organizational or location references

No Comments

No Known Activations