INDEX

Explanations

digits followed by the word "ometer"

oai_token-act-pair · gpt-3.5-turbo

terms related to measuring devices and their readings

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

lez

-0.74

Kut

-0.72

kus

-0.68

don

-0.66

ership

-0.65

giving

-0.65

paren

-0.62

hya

-0.62

ebin

-0.61

ABE

-0.61

POSITIVE LOGITS

ometer

1.00

ometers

0.81

ograph

0.80

ium

0.75

 gauge

0.73

 calibration

0.73

 measuring

0.72

asuring

0.68

 malfunction

0.68

ographs

0.68

Activations Density 0.030%

digits followed by the word "ometer"

terms related to measuring devices and their readings

No Comments

No Known Activations