INDEX

Explanations

phrases related to efficient operation and collaboration

oai_token-act-pair · gpt-3.5-turbo

phrases indicating effective collaboration or successful outcomes

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

Ħ¢

-0.74

ãĥīãĥ©

-0.68

myra

-0.68

grave

-0.68

lege

-0.68

 resurg

-0.67

 disbelief

-0.65

ngth

-0.63

artifacts

-0.63

raints

-0.62

POSITIVE LOGITS

ALLY

0.78

HQ

0.69

IFA

0.69

ework

0.65

MR

0.65

ooter

0.63

 hardest

0.63

umen

0.63

iously

0.63

 darn

0.63

Activations Density 0.126%

phrases related to efficient operation and collaboration

phrases indicating effective collaboration or successful outcomes

No Comments

No Known Activations