INDEX

Explanations

phrases related to various topics such as education, politics, crime, and environmental issues

oai_token-act-pair · gpt-3.5-turbo

phrases and terms related to recent events and processes

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

/-

-0.59

hani

-0.55

nces

-0.52

ATURES

-0.49

utenberg

-0.48

pixel

-0.48

aily

-0.48

kos

-0.47

 Caption

-0.46

-)

-0.45

POSITIVE LOGITS

 guise

0.59

 vicinity

0.51

 backdrop

0.50

esville

0.49

widget

0.49

 turbulent

0.47

 workplace

0.46

 unregulated

0.46

 foreseeable

0.45

 unfamiliar

0.44

Activations Density 1.427%

phrases related to various topics such as education, politics, crime, and environmental issues

phrases and terms related to recent events and processes

No Comments

No Known Activations